A redshifted codon-optimized firefly luciferase is a sensitive reporter for bioluminescence imaging

Abstract

Bioluminescence imaging has evolved as a powerful tool for monitoring biological processesin vivo. As transmission efficiency of light through tissue increases greatly for wavelengths above 600 nm we examined whether a redshifted codon-optimized firefly luciferase (λ_max = 615 nm) could be successfully employed as a sensitive reporter in mammalian cells. To this end, unmodified codon-optimized luciferase (λ_max = 557 nm) as well as the red-emitting S284T mutant luciferase were expressed simultaneously in human glioma cellsin vitro as well as in quadriceps muscles of mice in vivo. We show here that activity of the redshifted enzyme in human glioma cell culture approached approximately one-fourth of that seen with the unmodified enzyme. In contrast, light emission by the red-emitting luciferase in vivo was generally more efficient than that produced by its unmodified counterpart, most likely due to reduced absorption of red light by tissue. The mean ratio of light emission produced by the redshifted luciferase to that of the unmodified enzymein vivo was ∼3. Application of this new redshifted luciferase together with other optical reporters may be of considerable importance to biological research as it allows for imaging of deeper tissues as well as simultaneous monitoring of two molecular events in vitro and in vivo if appropriate filter sets are employed.

---

Introduction

In vivo imaging employing bioluminescence has emerged as a sensitive technique to monitor cellular processes such as tumor growth, metastasis, stem cell trafficking, or apoptosis in the same animal repetitively over time.¹Bioluminescence is based on the production of light by naturally occurring oxygenases such as luciferases. It typically yields low light intensities and - in contrast to fluorescence - is marked by the virtual absence of background and high sensitivity. No external excitation source is needed as is the case with fluorescent reporter proteins.

Luciferase from the North American firefly Photinus pyralis catalyzes the emission of yellow-green light (λ_max = 557 nm at 25 °C) from its substrate luciferin in the presence of ATP, Mg²⁺ and oxygen. The quantum yield of this reaction is 0.41² and sensitivity and range of linear response compare favorably with those of other reporter molecules such as β-galactosidase or green fluorescent protein.³ A sequence-optimized, so called “humanized” form of luciferase is the most commonly used reporter for non-invasive monitoring of gene expression in cell culture and small animals such as mice and rats.⁴ The sequence modifications introduced into wild-type luciferase by the manufacturer comprised adaptation to mammalian codon usage and removal of the C-terminal peroxisomal targeting sequence, transcription factor recognition sites, extended palindromes, restriction sites, and N-glycosylation sites. When expressed in mammalian cells this sequence-optimized, now cytosolically localized luciferase yielded 10- to over 100-fold greater activity than native luciferase and displayed a half-life (T_1/2) of 3 to 5 h, both factors depending on the cell line in use.⁴ For reasons of clarity, throughout this article the humanized luciferase optimized for mammalian expression will not be termed “wild-type” luciferase but will be referred to as “unmodified” luciferase as compared to the mutated redshifted humanized luciferase we describe in this article.

Bioluminescent light emission in vivo depends on the optical properties of tissues. It is reduced ∼10-fold per cm of tissue depth.⁵ Due to light scattering and absorption of mainly the shortwaved blue and green light by tissue it is generally assumed that only the light emitted above 600 nm can be efficiently detected by highly sensitive cooled charge-coupled device (CCD) cameras placed outside the animal.⁶ Blood hemoglobin and myoglobin which strongly absorb blue and green light are the main endogenous absorbers of light in vivo, but absorption is considerably less at wavelengths above 600 nm.⁷ Therefore, the development of redshifted reporters for bioluminescence imaging (BLI) seems promising and has been intensively pursued.^8–14

Using wild-type Photinus pyralis luciferase as the source material we have previously developed a mutant luciferase (S284T), in which serine at position 284 was exchanged by threonine, which led to a redshift of the emitted light (λ_max = 615 nm). When expressed in bacteria this mutant displayed a narrow emission band width, favorable kinetic properties, and good specific activity corresponding to 26% of the activity of wild-type luciferase.⁸

In the present study our objective was to integrate the S284T mutation into the humanized sequence-optimizedfirefly luciferase and validate this new reporter in mammalian cellsin vitro and in vivo. Luminometric analysis revealed that light production of the redshifted enzyme constituted approximately one-fourth of that of the unmodified enzyme in cell culture. However, in vivo mean light emission catalyzed by the redshifted enzyme after i.m. injection of the respective luciferase cDNAs was ∼3-fold stronger than that of unmodified luciferase, indicating that this new reporter may be particularly suitable for imaging of molecular events in deeper tissues.

Experimental

Materials

The human glioblastoma cell lines U87MG, U118, U138, U251, U343, U373, T98G, A172, LN18, LN229 and LN-Z308 were used for transient transfection of luciferase cDNA. Plasmid DNA for cell culture experiments was purified using the Qiagen plasmid Midi kit (Qiagen, Hilden, Germany). For animal experiments endotoxin-free plasmid DNA was prepared using the EndoFree plasmid Maxi kit (Qiagen).

Construction of luciferase expression plasmids

The cDNAs encoding either the humanized form of Photinus pyralis luciferase optimized with respect to its reporter function in mammalian cells⁴ (Promega, Madison, WI, USA) or its S284T mutant form were ligated into the expression vector pCDNA 3.1(−) (Invitrogen, Carlsbad, CA, USA). In these constructs luciferase expression was driven by the constitutively active human cytomegalovirus immediate-early (CMV) promoter.

Site directed mutagenesis

The serine to threonine exchange at amino acid position 284 (Ser284Thr) corresponding to a single nucleotide substitution was introduced into the luciferase plasmid described above using the QuikChange Site-directed Mutagenesis kit (Stratagene, La Jolla, CA, USA). The oligonucleotides used for mutagenesis were 5′-CAGGATTACAAGATTCAAACTGCGCTGCTGGTGCC-3′ and 5′-GGCACCAGCAGCGCAGTTTGAATCTTGTAATCCTG-3′. The entire luciferase cDNA was sequenced after mutagenesis and the substitution of nucleotide G (serine) by C (threonine) was confirmed.

Spectral analysis of luciferase light emission

The cDNAs encoding the humanized and sequence-optimized Photinus pyralis luciferase⁴ or its S284T mutant form, respectively, were ligated into the bacterial expression vector pQE30 (Qiagen) behind the T5 promoter and expressed in E. coli. After purification of the enzymes from E. coli lysates by affinity chromatography on Nickel-NTA agarose (Qiagen) their bioluminescence emission spectra were determined essentially as described previously.⁸

Transient transfection of human glioblastoma cells

Human glioma cells were liposomally transfected in parallel with 0.5, 1, and 2 μg of the two pCDNA3.1(−)-luciferase plasmids described above using either Transfectin (Bio-Rad, Hercules, CA, USA) or Lipofectamine (Invitrogen). Transfection protocols were first optimized for each cell line employing a plasmid containing enhanced green fluorescent protein (EGFP) cDNA expressed under control of the CMV promoter. Toxicity and transfection efficiency were assessed by fluorescence microscopy. Twelve to 24 hours before transfection cells were plated in 3.3 cm ø wells at a density of 70 to 80%. Transfection was performed either serum-free in Opti-MEM (Invitrogen) or in serum-containing medium according to the manufacturer's recommendations.

Luciferase activity assay

Cell lysates were prepared two days after the start of transfection. Protein content of the lysates was determined by the Bradford Protein Assay (Bio-Rad). Equal amounts of protein were analyzed luminometrically for luciferase activity with a Wallac Victor2 multilabel counter (Perkin Elmer, Waltham, MA, USA) with an integrated red-sensitive photomultiplier tube.¹⁵ All experiments were repeated at least twice. Light production was measured as “counts per second” (cps). Ratios of the activities of both enzymes for the respective cell line were calculated as mean values of all transfections performed and are reported along with the S.E.M.

Animal experiments

All animal procedures were performed in compliance with protocols that had been approved by the Animal Care and Use Committee at Martin Luther University Halle-Wittenberg.

Intramuscular plasmid injection

Endotoxin-free plasmid DNA harboring either the unmodified or the S284T mutated luciferase cDNA under control of the CMV promoter was dissolved in PBS at a concentration of 0.5 mg ml⁻¹. Three week old female Balb/c mice (n = 5) were anesthetized intraperitoneally (i.p.) with ketamine (75 mg/kg body weight) and xylazine (5 mg/kg body weight) and placed in a prone position with stretched hind legs. Both hind thighs were shaved and 100 μl of the respective plasmid solution were injected into both quadriceps muscles via syringe using a 30G needle inserted in a spacer tube to ensure a strictly identical injection depth of 2 mm. Three mice received the unmodified luciferase cDNA in the left and the S284T mutated luciferase cDNA in the right hind thigh while in the other two mice it was vice versa.

Bioluminescence imaging of mice

Bioluminescence imaging was done as previously described¹⁵ and started one day after plasmid injection i.m.. Mice were anesthetized as described above and injected i.p. with D-Luciferin (150 mg/kg body weight). Their hind thighs were shaved regularly to prevent light scattering. Animals were placed in a prone position in a dark box and light emission was recorded with a cooled CCD camera (VisiLuxx Imager, Visitron, Puchheim, Germany) starting 10 min after D-Luciferin administration. This camera has a quantum efficiency approaching 0.90 at wavelengths between 550 and 770 nm, indicating that one photon is converted to ∼0.9 electrons. Exposure time was 15 min using a pixel binning of 6. Light emission from both thighs was quantified as arbitrary light units using the Metamorph software (Visitron) as previously described.¹⁵ For each mouse ratios of photon emission from both thighs were calculated for all time points.

Results and discussion

Spectral analysis of luciferase light emission

Unmodified and S284T mutated sequence-optimized luciferases linked to a hexahistidine tag were expressed in E. coli, purified on Ni-NTA agarose, and subjected to spectral analysis and activity measurement. The humanized S284T enzyme had 22% specific activity compared to that of unmodified luciferase and catalyzed bioluminescence with an emission maximum of 616 nm at pH 7.8, corresponding to a 59 nm red shift in λ_max as compared to the unmodified enzyme. These values were essentially the same as those for the corresponding enzymes that had not been codon-optimized.⁸

We and others have previously shown that substitution of serine 284 by glycine, histidine, isoleucine, or asparagine also resulted in red-emitting luciferases (λ_max = 605–616 nm).^8,14 These mutants also displayed somewhat reduced activity, ranging from 7.9 to 60% as compared to wild-type luciferase, while their spectra all contained single emission maxima. Several other groups described single amino acid substitutions at the serine residue equivalent to position 284 in other firefly luciferases which all resulted in redshifted bioluminescence produced by the mutants.^9,11,16,17 Taken together, these studies support our findings that a single amino acid exchange in a firefly luciferase may result in a significant spectral shift of the emitted light.

Luciferase activities in human glioblastoma cells

Eleven human glioblastoma cell lines were transiently transfected with equal amounts (0.5, 1, and 2 μg) of either the unmodified or the S284T luciferase plasmid, which differed only in one nucleotide. In all cell lines, the unmodified luciferase demonstrated a clearly higher activity than the S284T mutant, irrespective of the amount of plasmid DNA administered (Fig. 1). The mean ratio of the activity of the unmodified luciferase to that of the S284T mutant luciferase in all cell lines was 4.28 (range 3.66 to 4.88), implying that the S284T luciferase displayed ∼23% activity of that of unmodified luciferase. This is well in line with the above reported results obtained after bacterial expression of these humanized luciferases, demonstrating that the S284T mutant luciferase was ∼22% active relative to the unmodified luciferase.

Fig. 1 Comparison of the enzymatic activities of unmodified Photinus pyralis luciferase and its S284T variant. Eleven human glioma cell lines were transiently transfected with equimolar amounts (0.5, 1, and 2 μg) of plasmid DNA encoding the respective luciferases. Cells transfected with equimolar amounts of DNA were analyzed luminometrically for bioluminescent activity which was determined as cps/μg protein. Mean ratios of the activity of unmodified luciferase to that of the S284T mutant luciferase are presented along with the S.E.M.

Bioluminescence imaging of mice

As red light penetrates tissue more deeply than yellow and green light we next asked whether the redshifted S284T mutant luciferase might be a more suitable and sensitive reporter for in vivo imaging. Gene therapy studies had proven that naked plasmid DNA can be efficiently internalized and long-term expressed by muscle cells of young mice in vivo without applying a special delivery strategy.^18,19 Injection of naked luciferase plasmid DNA into subcutaneous (s.c.) tumors resulted in considerably lower and more variable levels of reporter protein at the injection site as compared to intramuscular (i.m) injection.²⁰ Therefore, we chose the method of direct i.m. injection of naked plasmid DNA for in vivo comparison of the red-emitting luciferase with its unmodified counterpart. We directly injected plasmid DNAs harboring the cDNAs encoding humanized luciferase or its S284T variant into both thighs of young mice (n = 5) intramuscularly (i.m.) and compared light emission over time. Animals were imaged on days 1, 3, 5, 7, 12, and 18 after plasmid injection. On day 1, four mice already showed light emission from both thighs while in one mouse signals were first seen on day 3. In all mice bioluminescent signals from both thighs could be detected until day 18 (Fig. 2), confirming that a single reporter gene injection i.m. can lead to long-term expression of the transgene.¹⁹ One mouse demonstrated similar light emission from both thighs on day 1 and day 3 post plasmid injection, while on the subsequent time points, signals emitted from the thigh that had been injected with the S284T luciferase plasmid were clearly higher than signals from the contralateral thigh (Fig. 3). All other mice consistently displayed a higher light emission from the side that had been injected with the S284T luciferase plasmid compared to the contralateral side (Fig. 2 and 3). Mean light emission from the S284T luciferase cDNA-injected side on days 1, 3, 5, 7, 12, and 18 post plasmid injection was 2.6, 2.9, 2.3, 2.8, 3.3, and 3.5-fold higher than photon emission from the contralateral side, respectively. As the S284T mutated luciferase had shown clearly less activity in vitro as compared to the unmodified enzyme, we conclude that the spectral redshift observed with this mutant allows for considerably better transmission of light through tissue in vivo.

Fig. 2 Bioluminescence imaging of two mice over time. The animals had been injected i.m. on day 0 with plasmid DNAs carrying either the unmodified (left hind thigh) or the S284T mutant (right hind thigh) luciferase cDNA. Light production from the right thighs of both mice (placed in prone position) was consistently stronger.

Fig. 3 Comparison of the light emission from both thighs of mice (n = 5) that had been injected i.m. with plasmid DNAs carrying either unmodified or S284T mutant luciferase cDNA. Depicted are the ratios of light production from the quadriceps muscle that had been injected with S284T mutant luciferase cDNA to light emission from the contralateral thigh that had been injected with unmodified luciferase cDNA. Images were taken on days 1, 3, 5, 7, 12, and 18 post injection.

Comparing light emission over time two mice demonstrated peak signals from both thighs on day 5 while in the other mice peak signals for unmodified and S284T luciferase were seen on days 5 and 12, days 7 and 12, and days 12 and 18, respectively. Our data agreed well with previously published studies, demonstrating that onset of gene expression after cDNA injection i.m. is considerably slower than in other tissues, e.g. in liver.¹⁸ In several studies expression from plasmid DNA vectors injected i.m. peaked at 14 days post plasmid injection or even later.^18,21,22 The fact that in three mice peak activity seen with the S284T mutant luciferase occurred somewhat later than peak activity observed for unmodified luciferase may indicate increased stability of the mutant enzyme as compared to the unmodified enzyme. The highest photon emission measured throughout the study at a site that had been injected with S284T luciferase cDNA was ∼3.2-fold higher than the highest signal detected at a site that had been injected with the unmodified luciferase cDNA. Taken together, our data suggest that the S284T mutant luciferase was superior to the commonly used unmodified luciferase reporter in monitoring gene expressionin vivo and might therefore be preferentially used for imaging of biological processes occurring in deeper tissues.

Regarding wild-type Photinus pyralis luciferase and its S284T mutant form, we showed previously that the purified enzymes have a relatively short half-life (T_1/2) of less than 1 hour when expressed in bacteria.¹² Interestingly, while purified wild-type luciferase had a T_1/2 of 15.6 min, introduction of the S284T mutation not only shifted the spectrum to the red but also increased T_1/2 to 48 min. Notably, due to the method selected for enzyme purification both luciferases still carried an additional N-terminal tag of 5 amino acids¹² which also may have had some impact on enzyme stability.²³ Firefly luciferases are known to be more stable when expressed in mammalian cells (T_1/2 = 3–5 h),^24,25 presumably due to protection by chaperones.²⁶ It is conceivable that introduction of the S284T mutation into the sequence-optimized luciferase we employed might also have stabilized the enzyme to some degree, which, in turn, may have contributed to the stronger light emission we observed in vivo for the S284T mutant as compared to the unmodified enzyme. However, it is unlikely that higher stability of the mutant enzyme should be the main cause for the higher light production we measured in vivo as a longer T_1/2 of the mutant enzyme should also be reflected by increased enzyme activity in cell culture.

Our in vivo study demonstrated that in general light emission catalyzed by the redshifted mutant S284T luciferase was considerably higher than that observed for unmodified luciferase, although activity ratios differed between the animals (Fig. 3). Wolff et al. reported substantial variations in luciferase activity after direct injection of luciferase plasmid DNA into the quadriceps muscle of 6 week old mice.²⁷ The standard error was generally around one-third to one-half of the mean. It was shown that this was not caused by variability in injection technique but rather by unequal degradation of plasmid DNA.²⁷ Variable plasmid stability in different mice may also contribute to differences in light emission found in our study.

An alternative approach for comparing light production catalyzed by different luciferases would have been to inject mice s.c. with tumor cells stably transfected with the respective luciferase cDNAs. We chose to directly inject the luciferase cDNAs i.m. for several reasons: different growth rates of the stable cell clones leading to different light production at a given time point might be misinterpreted as a difference in specific luciferase activity. In addition, hypoxia often seen during tumor growth might result in diminished luciferase activity, as oxygen is essential for light production.²⁸ Thus, reduced light emission from a tumor due to hypoxia might falsely be taken as reduced specific activity of one of the luciferases.

Dual reporter imaging in vivo has also been previously reported by Bhaumik et al., who compared Photinus pyralis luciferase with luciferase from the seapansy Renilla reniformis, which emits blue-green light (λ_max = 480 nm) and uses coelenterazine as substrate.²⁹ However, the broad use of this enzyme as a reporter for serial in vivo monitoring has been seriously hampered by the strong tissue absorbance of blue-green light and autoluminescence, high clearance rate and expense of the substrate.³⁰ Therefore, the development and examination of redshifted firefly luciferases for dual reporter assays seems a promising approach.

Recently, a thermostabilized form of the S284T mutant luciferase¹² has been employed in multicolor reporter assays using lysates prepared from human hepatoblastoma cells.³¹ Unlike the S284T mutant we used here this luciferase had not been codon-optimized for use in mammalian cells and still contained the peroxisomal targeting sequence as well as other interfering sequences,⁴ thus rendering this reporter rather suboptimal for gene expression studies in eukaryotic cells.⁴ We suggest that a thermostabilized variant of the codon-optimized and cytosolically localized redshifted S284T mutant luciferase described here may constitute an even more powerful reporter for multianalyte in vitro assays as well as in vivobioluminescence imaging.

Conclusion

We show here for the first time that the redshifted codon-optimized S284T luciferase mutant is a sensitive tool for monitoring gene expressionin vivo, suggesting that it might be more suitable for imaging of biological processes occurring in deeper tissues as compared to unmodified luciferase. Further improvement of reporter sensitivity may be achieved by integrating thermostabilizing mutations into this redshifted enzyme.¹²

The development of a panel of luciferases with different spectral properties will not only allow for multiplex reporter assays in vitro but may also facilitate the simultaneous monitoring of several biological processes in the same animal in vivo, if a set of appropriate filters is applied. In addition, the detection of photon emission from deeper tissues may allow for threedimensional image reconstruction and thus a more precise determination of the localization and size of a light source.³²

Acknowledgements

Support by the National Science Foundation (grant MCB 0444577), the Air Force Office of Scientific Research (grant FA9550-07-1-0043) and the foundations Wilhelm Sander-Stiftung (grant 2004.075.1) and Else Kröner-Fresenius-Stiftung (grant P35/04/A45/04) is gratefully acknowledged. We thank Prof. Winfried Burkert for continuous encouragement.

References

1. H. Masuda, H. J. Okano, T. Maruyama, Y. Yoshimura, H. Okano and Y. Matsuzaki, In vivo imaging in humanized mice, Curr. Top. Microbiol. Immunol., 2008, 324, 179–96.
2. Y. Ando, K. Niwa, N. Yamada, T. Enomoto, T. Irie, H. Kubota, Y. Ohmiya and H. Akiyama, Firefly bioluminescence quantum yield and colour change by pH-sensitive green emission, Nat. Photonics, 2008, 2, 44–47.
3. L. H. Naylor, Reporter gene technology: the future looks bright, Biochem. Pharmacol., 1999, 58, 749–57.
4. D. J. Groskreutz, B. A. Sherf, K. V. Wood and E. T. Schenborn, Increased expression and convenience with the new pGL3 luciferase reporter vectors, Promega Notes, 1995, 50, 2–6.
5. C. H. Contag, P. R. Contag, J. I. Mullins, S. D. Spilman, D. K. Stevenson and D. A. Benaron, Photonic detection of bacterial pathogens in living hosts, Mol. Microbiol., 1995, 18, 593–603.
6. H. Zhao, T. C. Doyle, O. Coquoz, F. Kalish, B. W. Rice and C. H. Contag, Emission spectra of bioluminescent reporters and interaction with mammalian tissue determine the sensitivity of detection in vivo, J. Biomed. Opt., 2005, 10, 41210.
7. B. W. Rice, M. D. Cable and M. B. Nelson, In vivo imaging of light-emitting probes, J. Biomed. Opt., 2001, 6, 432–40.
8. B. R. Branchini, T. L. Southworth, N. F. Khattak, E. Michelini and A. Roda, Red- and green-emitting firefly luciferase mutants for bioluminescent reporter applications, Anal. Biochem., 2005, 345, 140–8.
9. N. Kajiyama and E. Nakano, Isolation and characterization of mutants of firefly luciferase which produce different colors of light, Protein Eng., 1991, 4, 691–3.
10. B. F. Eames, D. A. Benaron, D. K. Stevenson and C. H. Contag, Construction and characterization of a red-emitting luciferase, SPIE, 1999, 3600, 36–39.
11. T. Arslan, S. V. Mamaev, N. V. Mamaeva and S. M. Hecht, Structurally modified firefly luciferase. Effects of amino acid substitution at position 286, J. Am. Chem. Soc., 1997, 119, 10877–87.
12. B. R. Branchini, D. M. Ablamsky, M. H. Murtiashaw, L. Uzasci, H. Fraga and T. L. Southworth, Thermostable red and green light-producing firefly luciferase mutants for bioluminescent reporter applications, Anal. Biochem., 2007, 361, 253–62.
13. N. K. Tafreshi, S. Hosseinkhani, M. Sadeghizadeh, M. Sadeghi, B. Ranjbar and H. Naderi-Manesh, The influence of insertion of a critical residue (Arg356) in structure and bioluminescence spectra of firefly luciferase, J. Biol. Chem., 2007, 282, 8641–7.
14. E. Shapiro, C. Lu and F. Baneyx, A set of multicolored Photinus pyralis luciferase mutants for in vivo bioluminescence applications, Protein Eng. Des. Sel., 2005, 18, 581–7.
15. A. Söling, C. Theiss, S. Jungmichel and N. G. Rainov, A dual function fusion protein of Herpes simplex virus type 1 thymidine kinase and firefly luciferase for noninvasive in vivo imaging of gene therapy in malignant glioma, Genet. Vaccines Ther., 2004, 2, 7–21.
16. N. K. Tafreshi, M. Sadeghizadeh, R. Emamzadeh, B. Ranjbar, H. Naderi-Manesh and S. Hosseinkhani, Site-directed mutagenesis of firefly luciferase: implication of conserved residue(s) in bioluminescence emission spectra among firefly luciferases, Biochem. J., 2008, 412, 27–33.
17. S. V. Mamaev, A. L. Laikhter, T. Arslan and S. M. Hecht, Firefly luciferase: alterations of the color of emitted light resulting from substitutions at position 286, J. Am. Chem. Soc., 1996, 118, 7243–4.
18. J. A. Wolff and V. Budker, The mechanism of naked DNA uptake and expression, Adv. Genet., 2005, 54, 3–20.
19. J. A. Wolff, J. J. Ludtke, G. Acsadi, P. Williams and A. Jani, Long-term persistence of plasmid DNA and foreign gene expression in mouse muscle, Hum. Mol. Genet., 1992, 1, 363–9.
20. A. Kawase, T. Nomura, T. K. Yasuda, N. Kobayashi, M. Hashida and Y. Takakura, Disposition and gene expression characteristics in solid tumors and skeletal muscle after direct injection of naked plasmid DNA in mice, J. Pharm. Sci., 2003, 92, 1295–304.
21. H. L. Davis, R. G. Whalen and B. A. Demeneix, Direct gene transfer into skeletal muscle in vivo: factors affecting efficiency of transfer and stability of expression, Hum. Gene Ther., 1993, 4, 151–9.
22. M. Manthorpe, F. Cornefert-Jensen, J. Hartikka, J. Felgner, A. Rundell, M. Margalith and V. Dwarki, Gene therapy by intramuscular injection of plasmid DNA: studies on firefly luciferase gene expression in mice, Hum. Gene Ther., 1993, 4, 419–31.
23. A. Bachmair, D. Finley and A. Varshavsky, In vivo half-life of a protein is a function of its amino-terminal residue, Science, 1986, 234, 179–86.
24. J. F. Thompson, L. S. Hayes and D. B. Lloyd, Modulation of firefly luciferase stability and impact on studies of gene regulation, Gene, 1991, 103, 171–7.
25. G. M. Leclerc, F. R. Boockfor, W. J. Faught and L. S. Frawley, Development of a destabilized firefly luciferase enzyme for measurement of gene expression, Biotechniques, 2000, 29, 590–8.
26. J. E. Souren, F. A. Wiegant and R. Van Wijk, The role of hsp70 in protection and repair of luciferase activity in vivo; experimental data and mathematical modelling, Cell Mol. Life Sci., 1999, 55, 799–811.
27. J. A. Wolff, P. Williams, G. Acsadi, S. Jiao, A. Jani and W. Chong, Conditions affecting direct gene transfer into rodent muscle in vivo, Biotechniques, 1991, 11, 474–85.
28. E. H. Moriyama, M. J. Niedre, M. T. Jarvi, J. D. Mocanu, Y. Moriyama, P. Subarsky, B. Li, L. D. Lilge and B. C. Wilson, The influence of hypoxia on bioluminescence in luciferase-transfected gliosarcoma tumor cells in vitro, Photochem. Photobiol. Sci., 2008, 7, 675–80.
29. S. Bhaumik, X. Z. Lewis and S. S. Gambhir, Optical imaging of Renilla luciferase, synthetic Renilla luciferase, and firefly luciferase reporter gene expression in living mice, J. Biomed. Opt., 2004, 9, 578–86.
30. M. Otto-Duessel, V. Khankaldyyan, I. Gonzalez-Gomez, M. C. Jensen, W. E. Laug and M. Rosol, In vivo testing of Renilla luciferase substrate analogs in an orthotopic murine model of human glioblastoma, Mol. Imaging, 2006, 5, 57–64.
31. E. Michelini, L. Cevenini, L. Mezzanotte, D. Ablamsky, T. Southworth, B. R. Branchini and A. Roda, Combining intracellular and secreted bioluminescent reporter proteins for multicolor cell-based assays, Photochem. Photobiol. Sci., 2008, 7, 212–7.
32. H. Dehghani, S. C. Davis, S. Jiang, B. W. Pogue, K. D. Paulsen and M. S. Patterson, Spectrally resolved bioluminescence optical tomography, Opt. Lett., 2006, 31, 365–7.

---