Development and Characterization of a Novel CD19CherryLuciferase (CD19CL) Transgenic Mouse for the Preclinical Study of B-Cell Lymphomas

Presently, the preclinical in vivo assessment of different therapeutic interventions is determined using xenograft implantation of human tumor cell lines in immunocompromised mice (1). These models, although obviously allowing for direct measurement of tumor in animals bearing the xenograft do not necessarily recapitulate the natural history of the disease, especially when it comes to assessing malignancies of the hematopoietic system. For example, the presently used xenograft models fail to integrate the impact of (i) the appropriate tumor microenvironment, especially when a disease such as lymphoma is implanted in the subcutaneous environment; (ii) an intact immune system on the effects of the tumor as a function of the intervention; (iii) bone marrow involvement on outcome; (iv) therapeutic agents on more realistic tumor volumes at time 0, as tumor xenograft experiments are typically conducted on small volume tumors to comply with present IAUC guidelines. Genetically engineered mouse models (GEMM) of cancer and lymphoma, in particular, circumnavigate many of these liabilities and offer the opportunity to investigate the efficacy of anticancer agents in the context of the correct tumor microenvironment, tumor volume, and an intact immune system (2, 3).

One major barrier to the use of GEMM that develop spontaneous disease revolves around the ability to quantitatively assess tumor growth as a function of a discrete intervention. In the absence of an in vivo imaging system, such experiments would require a large number of animals to be sacrificed at multiple time points to overcome variability between animals, as well as the variability related to the intervention (4). Over the past several years, the utility and sophistication of mouse imaging technologies has grown substantially and is increasingly playing a critical role in the field (5, 6). In particular, in vivo bioluminescent imaging (BLI) offers the advantage of longitudinal assessments in which multiple measurements of tumor growth can be collected for a given animal, reducing the requirement for large numbers of mice in preclinical experiments (7–9). The firefly luciferase from Photinus pyralis has been commonly used as a bioluminescent marker in biomedical research as it provides excellent sensitivity for optical detection. Recently, new developments in in vivo imaging include the use of dual-functioning reporters that are both fluorescent and bioluminescent, providing maximum experimental flexibility that enables unique biologic applications (10).

We report the development of a novel transgenic mouse using a CherryLuciferase fusion protein and the CD19 as the target locus to achieve functional imaging of B-cell lymphomagenesis in vivo. The use of a dual function protein enables one to link the in vivo analysis via BLI to cell discriminating assessment ex vivo using confocal microscopy. By restricting the fluorescent–bioluminescent emission in genetically engineered mouse models of living mice to the B-cell lineage, we have generated a murine model of lymphoma that allows for the efficient and faster assessment of new drugs and new drug combinations. This model can be used to hasten innovative discovery and development of novel therapeutic strategies in lymphoma.

All mice were housed in a barrier, pathogen-free mouse facility at the animal facility. All animal studies were carried out in accordance with institutional guidelines and an approved IACUC protocol.

The Cherry coding sequence without the STOP codon was obtained by PCR amplification using pmCherry (from Dr. Roger Y. Tsien, Department of Chemistry and Biochemistry, University of San Diego) as the template and primers containing engineered HindIII restriction sites for cloning. The primers used were as follows: 5′-TTTTAAGCTTCCACCATGGTGAGCAAGG-3′ and 5′-TTTTAAGCTTCTTTCTTGTACAGCTCGTCC-3′. After HindIII restriction, the purified PCR product was cloned into the HindIII site of pGL4.13[luc2/SV40] plasmid (11) bearing the synthetic firefly luciferase gene under the control of the SV40 early enhancer/promoter, thus obtaining the pGLCherryLuciferase plasmid, where the cherry and the luciferase genes formed one open reading frame. The dual bioluminescent/fluorescent activity of the fusion protein was evaluated by flow cytometry using a FACSCalibur (Becton Dickinson) and a luciferase assay (Promega Corporation) of pGLCherryLuciferase-transfected human embryonic kidney 293 (HEK 293) cells.

The targeting vector (a gift from Dr. Riccardo Dalla Favera, Institute of Cancer Genetics, Columbia University, NY) has already been described (12), with the exception that the homology of the short arm was increased by substituting the 300-bp fragment upstream of exon 1 with a 1,741-bp fragment obtained by PCR amplification of embryonic stem (E14-1) cell DNA using oligonucleotides tagged with restriction sites at their 5′ end. The following primers were used: 5′-TTTGCGGCCGCGTGCATACACACTGTACCACCTC/TAGATGCATCCCTGGGGCACCCGGCTTCTGCGTG-3′. Finally the substitution of the cre/neo cassette with a CherryLuciferase/neo cassette with SalI restriction sites at each end of the cassette generated the CD19CherryLuciferase targeting vector used to transfect mouse embryonic stem cells. Transfection of Not I linearized vector of embryonic stem (ES1-14) cells was followed by screening of 200 embryonic stem cell colonies by PCR amplification of embryonic stem extract genomic DNA using transgene and locus-specific primers (Supplementary Table S1). One positive clone was injected into C57BL/6J blastocysts and 2 chimeric mice were obtained. The chimeric mice were genotyped by PCR for the presence of both wild-type and knockin alleles and one of the chimera transmitted the targeted CD19 locus to subsequent generations (Supplementary Fig. S1). The neo cassette was deleted by crossing the resulting mice to the loxp deleter Rosa26Cre (a gift from Dr. Antonio Iavarone, Institute of Cancer Genetics, Columbia University, NY). The resulting CD19CherryLuciferase line was then backcrossed to C57BL/6 mice. Transfection of embryonic stem cells and C57BL/6J blastocysts injection were carried out at the transgenic mouse facility of the Herbert Irving Comprehensive Cancer Center, Columbia University, NY.

In vivo BLI was conducted on a cryogenically cooled IVIS system (Xenogen Corp.). Ten to 15 minutes before in vivo imaging animals were injected intraperitoneally with a saline solution of 15 mg/mL of d-luciferin (Caliper Life Sciences) at a dose of 150 mg/kg body weight and 5 minutes later anesthetized with a mixture of 2% isofluorane/air. During image acquisition, isofluorane anesthesia was maintained using a cone delivery system integrated with the IVIS system. A gray scale body surface image was collected in the chamber under dim illumination followed by acquisition and overlay of the pseudocolor image representing the spatial distribution of detected photon counts emerging from active luciferase within the animal. Images and measurements of bioluminescent signals were acquired and analyzed using Living Image_ software (Xenogen). Image data are displayed in photons/sec/cm²/sr. Camera settings as integration time, binning, f/stop, and field of view were kept constant during all measurements. Mice were initially imaged at the weaning stage and thereafter imaged weekly. To be eligible for inclusion in a preclinical trial cohort, mice must have a detectable bioluminescence signal, and the intensity of the bioluminescence signal must increase between days 1 and 7. We consider a good starting point a signal between 2 × 10⁷ and 6 × 10⁷ (tumor weight of 100–200 mg). Mice are then randomized into groups by assigning them in order of intensity to the treatment groups to make sure that the average starting signals within all groups are as close as possible. No less than 5 mice were assigned to each group. All the mice are monitored twice a week and terminated if they become ill (lack normal grooming and avoidance behaviors), are unable to eat or drink, or if the tumor becomes overly large as to hinder normal body movement as by IACUC regulations.

For the in vivo bioluminescence analysis of CD19CherryLuciferase and double transgenic mice, we evaluate whether a statistically significant association existed between mean signal strength on days 1, 7, 14, and 21 in both the control (single transgenic mice) and experimental (double transgenic mice) groups. We carried out a repeated measures ANOVA within both the control and experimental groups, and the difference in mean values of signal expression was not significant in the control group (P = .13) but highly significant in the experimental group (P < .0001). For the evaluation of the effect of dexamethasone on tumor growth, we evaluated whether a statistically significant association existed between mean signal strength on days 0, 1, and 2 in both the control (untreated mice) and experimental (treated mice) groups. We carried out a repeated measures analysis within both the control and experimental groups and the difference in mean values of signal expression was significant in the control (P < .0001) and experimental groups (P < .014). Statistical analyses were carried out using SAS version 9.2 (SAS Institute).

At autopsy selected tissues were fixed in 10% formalin (Sigma) and prepared for standard histopathology evaluation. For confocal analysis, tissues were placed in 4% paraformaldehyde in PBS (pH 7.4) for 15 minutes. Washed in PBS and incubated overnight in 30% sucrose at 4°C. The next day the tissues were transferred in a solution of freezing medium and 30% sucrose (ratio 1:1) and incubated overnight at 4°C. The next day the tissues were transferred in 100% freezing medium and stored at −20°C. The cut sections were allowed to dry for 20 minutes before processing and stained with CD19. In brief, sections were washed in PBS (pH7.4) for 5 minutes, PBS-T (PBS with Trition 100, 0.1%) for 5 minutes and blocked for 1 hour in PBS-T with 5% normal goat serum (Invitrogen Corporation). Primary antibody purified rat anti-mouse CD19 (BD Pharmigen) was diluted 1:500 in PBS and left overnight at room temperature. The next day the sections were washed in PBS 3 times for 15 minutes and appropriate secondary antibody Alexa 488 (Invitrogen Corporation) at 1:250 dilution was applied to the sections for 1 hour. Sections were then mounted using mounting media with 4′, 6-diamidino-2-phenylindole (DAPI) from Vectashield (Vector Laboratories). Images were collected on a Leica TCS SP 5 laser scanning confocal microscope.

The luciferase activity of lymphomas in the double transgenic mice was measured at various stages of tumor development. Animals were imaged, sacrificed, and lymph node tumor mass was isolated and the signals were plotted against the wet tumor weight. Linear correlation between detected photons/sec/cm2/sr measurements and tumor weight was estimated by the calculation of the Pearson product-moment correlation coefficient.

We constructed a fusion protein consisting of the monomeric mutant red fluorescent Cherry (10) and the synthetic firefly Luciferase by cloning the Cherry gene into the plasmid vector pGL4.13[luc2/SV40] carrying the luciferase gene (Fig. 1A). The pGLCherryLuciferase expression vector was then transduced into HEK 293 cells by calcium phosphate–mediated transfection. Twenty-four and 48 hours after transfection, cells were harvested and analyzed by flow cytometry and luciferase activity (Fig. 1B and C). The results indicated that the chimeric protein was efficiently expressed in HEK 293 cells and that both proteins retained their functional properties in vitro. To express the CherryLuciferase fusion protein in cells of the B lineage in mice, the CD19 locus was targeted. CD19 is expressed in cells of the B lineage in the early stages and throughout B-cell development and differentiation, but it is lost on maturation to plasma cells (13). To obtain CD19-directed CherryLuciferase expression, we have used a modified version of the CD19-cre targeting vector (12). Briefly the cre/neo gene cassette was substituted by a CherryLuciferase/neo cassette and the 330-bp fragment upstream of exon1 by a 1,741-bp fragment obtained by PCR amplification of embryonic stem (E14-1) cell DNA using oligonucleotides tagged with restriction sites at their 5′ end (Fig. 1D). Mice hemizygous for the CherryLuciferase insertion (designated CD19CherryLuciferase mice), which retain one functional CD19 allele, are phenotypically normal and did not display alterations of thymus, spleen, or lymph nodes or develop tumors, as shown by gross and histologic examinations (Supplementary Fig. S2). Heterozygous mice are also similar to their wild-type littermates with respect to number of B cells in peripheral lymphoid tissues (Data not shown).

B cells originate in the bone marrow and enter the blood stream and migrate to the peripheral lymphoid organs where they complete their maturation. Detection of the chimeric CherryLuciferase protein expression via in vivo luciferase imaging, 10 minutes after intraperitoneal injection of luciferin into the anesthetized animals, showed that luciferase activity is retained by the CherryLuciferase fusion protein in vivo. Bioluminescence was restricted to regions where spleen and lymph nodes are known, which is consistent with the B-cell–restricted expression of the fusion protein (Fig. 2A). Moreover intraperitoneal injection of d-luciferin into control littermates showed no luciferase activity. We concluded that the high sensitivity of bioluminescence imaging allows for the noninvasive quantification of luciferase expression in the secondary lymphoid organs of the CD19CherryLuciferase mice.

To assess whether Cherry fluorescent protein is coexpressed with the CD19 antigen, mouse spleen sections of a CD19CherryLuciferase transgenic mouse were labeled with rat anti–mouse-CD19 antibody and examined by confocal microscopy for the expression of CD19 and Cherry (Fig. 2B). Spectral analysis of the fluorescent signal clearly identified Cherry fluorescence and of the CD19 in the CD19CherryLuciferase mouse spleen sections confirming the expected heterozygosity of the CD19 locus. Superposition of the fluorescence for Cherry with CD19 clearly showed colocalization of the 2 proteins in B cells, confirming that the transcription of the CherryLuciferase fusion protein is under the control of the CD19 locus regulatory elements.

Cherry fluorescence was also identified in peripheral blood mononuclear cell (PBMC) smears of transgenic CD19CherryLuciferase mice when compared with equivalent PBMC smears of control littermates. Moreover, luciferase activity was detected in cell extracts of CD19-positive cells purified from PBMCs of transgenic CD19CherryLuciferase mice when compared with equivalent cell extracts of control littermates (Supplementary Fig. S3).

The CD19CherryLuciferase transgenic mouse was generated with the intent to facilitate the use of GEMM of lymphomas in longitudinal studies. To validate its application, we crossed the CD19CherryLuciferase with a GEMM of an aggressive B-cell lymphoma (14, 15), a cancer associated with a chromosomal translocation of the c-myc gene. λ-MYC heterozygous transgenic animals die between days 38 and 216 with advanced lymphadenopathy but minimal splenomegaly (14, 15). Double transgenic mice were produced by breeding λ-MYC heterozygous transgenic animals with CD19CherryLuciferase heterozygous transgenic animals. As expected, around 25% of the progeny inherited both transgenes. After weaning, double transgenic animals were closely inspected at weekly intervals and imaged every month to detect any abnormal signal indicative of lymph nodes enlargement. Mice suspected of developing lymphoma were examined more frequently and imaged weekly. The spatiotemporal tracking of a B-cell lymphoma growth in double transgenic animals was compared with single transgenic animals, during a 3 weeks time period, and is shown in Fig 3A. A mediastinal bioluminescent signal was initially detected by in vivo imaging analysis of the double transgenic animals. The intensity of the bioluminescent signal clearly increased as the tumor grew with subsequent involvement of inguinal and cervical lymph nodes and wide dissemination in the final stage of the disease. On the contrary, the intensity of the bioluminescent signal in the single transgenic animals remained unchanged and confined to cervical lymph nodes and spleen (Fig. 3A and B). Histologic examination of affected animals confirmed that the tumors originated in the lymph nodes, as had been previously described (14), with involvement of the spleen but only in the late stages of the disease (Fig. 4). To evaluate the response of an established B-cell lymphoma to a drug known to induce remissions, double transgenic mice with diffuse disease were treated with dexamethasone (2 doses of 4 mg/kg) administrated by intraperitoneal injection. A substantial reduction in signal intensity, corresponding to a decrease in tumor burden, was observed within 24 hours of the first intraperitoneal injection followed by a more moderated response to the second dexamethasone administration. The response to the drug treatment was similar in all mice with an established B-cell lymphoma, with no decrease in signal intensity being observed in the untreated mice (Fig. 5). These findings validated the use of the CD19CherryLuciferase transgenic mouse in the assessment of an active drug routinely used in the treatment of lymphoproliferative malignancies.

Xenograft models have become the major tool for assessing the preclinical activity of new drugs and drug: drug combinations before early-phase clinical study. Although these models provide important information about cause and effect, they possess, as discussed above, a number of different liabilities. In the case of lymphoma, they do not reflect the proper localization of the disease to the appropriate in vivo compartment, making conclusions about effect tenuous. One major criticism of these models is that they do not accurately predict what happens in the clinical setting. Although that may be true, it is also clear that there are a number of factors unrelated to the actual murine model that may account for this. In particular, preclinical murine models are intended to function as tools, used to refine scientific and pharmacologic principles before study in humans. However, in many cases, schedules and doses identified in preclinical models as active are modified to alternative and more convenient schedules in the clinic. Thus, the opinion that preclinical models do not accurately predict the probability of success in the clinic is complicated by a number of both murine-dependent and murine-independent factors.

At present there are no other models that allow in vivo imaging of spontaneously generated lymphoma. The concept of generating a single fusion gene under the control of an intrinsic B-cell–restricted promoter, CD19, allows both elements of the fusion gene to be expressed simultaneously and to the same extent. The expression of the luciferase protein in CD19-restricted cells allows for the imaging of B cells in lymphatic tissue without having to sacrifice the animal. Expression of the cherry fluorescent protein affords the ability to use flow cytometry and confocal microscopy to image the trafficking and migration of CD19-positive B cells into different tissue and organ compartments as a function of different stimuli. Moreover, the presence of an intact CD19 allele assures the evaluation of tumor in the context of a functioning immune system. Studies aimed at elucidating the function of CD19 in vivo by generating CD19-deficient mice have shown that ablation of the CD19 locus leads to a significant reduction in the number of B cells in peripheral lymphoid tissues and a markedly decreased proliferative response to mitogens. However, heterozygous mice have been shown to be similar to their wild-type littermates with respect to number of B cells and their proliferative response (16).

One of the more flexible aspects of the model is retained in the fact that the CD19 CherryLuciferase mouse can be crossed with most other models that spontaneously develop lymphoma. Presently, there are a number of mice that spontaneously generate various forms of B-cell non-Hodgkin lymphoma (NHL), though most of these models seem to recapitulate the rare and rapidly growing Burkitt lymphoma. The 2008 World Health Organization (17) now recognizes nearly 65 different subtypes of NHL, very few of which have been recapitulated in a GEMM. One of the strengths of the CD19CherryLuciferase murine model is that it can be crossed with other spontaneous murine B-cell lymphoma models allowing for the study B-cell trafficking and or response to therapy in a “disease” or lymphoma subtype-specific manner. The use of noninvasive approaches in imaging tumor development in small animals is rapidly expanding, and within this context, BLI represents a powerful noninvasive tool for longitudinal study of tumor growth in mouse.

The validity of noninvasive quantification of the BLI signal as a parameter for tumor load has been tested in various methods, and in all cases, good correlations were found between BLI quantifications and tumor volume or tumor weight (18–21) We have also quantified the bioluminescence signals detectable, sacrificed the double transgenic mice, harvested their lymph nodes derived tumors, and correlated tumor weights with the bioluminescence signals from region of interest (ROI). We showed that there was a linear correlation between tumor weights and bioluminescence light emission, indicating that bioluminescence imaging is an appropriate surrogate means of assessing tumor growth throughout the course of preclinical trials (Fig. 6).

The flexibility to image different mouse models of lymphoma within the same methodology and the ability to assess the efficacy of new drug candidates in different GEMMs are 2 of the advantages presented by the use of the CD19CherryLuciferase knockin transgenic mouse. It is widely accepted that GEMMs possess many of the desirable features of a mouse model with regard to drug development and that they will become an obligatory tool in validating and prioritizing new compounds. The mouse model described here is intended to be used to hasten translational studies of novel agents in lymphoma, with the intent that understanding the relevant pharmacology before clinical study will accelerate successful development in clinical studies.

No potential conflicts of interest were disclosed.

Conception and design: L. Scotto, O.A. O'Connor

Development of methodology: L. Scotto, M. Kruithof-de Julio, O.A. O'Connor

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): L. Scotto, M. Kruithof-de Julio, L. Paoluzzi, M. Kalac, E. Marchi, J.B. Buitrago, J. Amengual, M.M. Shen, O.A. O'Connor

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): L. Scotto, M. Kruithof-de Julio, L. Paoluzzi, M. Kalac, E. Marchi, J. Amengual, O.A. O'Connor

Writing, review, and/or revision of the manuscript: L. Scotto, M. Kruithof-de Julio, M. Kalac, O.A. O'Connor

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): L. Scotto, M. Kruithof-de Julio, J.B. Buitrago, O.A. O'Connor

Study supervision: L. Scotto, O.A. O'Connor

The authors thank Sean Clark-Garvey for statistical analysis; R. Dalla Favera and A. Iavarone for providing reagents; and R. Dalla Favera for critical review.

M.M. Shen and M. Kruithof-de Julio are supported by NIH grants (DK076602 and CA084294).

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.