Thermostability of Firefly Luciferases Affects Efficiency

RESEARCH ARTICLE

Molecular Imaging . Vol. 3, No. 4, October 2004, pp. 324 – 332

324

Thermostability of Firef ly Luciferases Affects Efficiency of Detection by In Vivo Bioluminescence Brenda Baggett 1*, Rupali Roy 1*, Shafinaz Momen1, Sherif Morgan1, Laurence Tisi 2, David Morse1, and Robert J. Gillies1 1

University of Arizona and 2Cambridge University

Abstract Luciferase from the North American firefly (Photinis pyralis) is a useful reporter gene in vivo, allowing noninvasive imaging of tumor growth, metastasis, gene transfer, drug treatment, and gene expression. Luciferase is heat labile with an in vitro halflife of approximately 3 min at 37°C. We have characterized wild type and six thermostabilized mutant luciferases. In vitro, mutants showed half-lives between 2- and 25-fold higher than wild type. Luciferase transfected mammalian cells were used to determine in vivo half-lives following cycloheximide inhibition of de novo protein synthesis. This showed increased in vivo thermostability in both wild-type and mutant luciferases. This may be due to a variety of factors, including chaperone activity, as steady-state luciferase levels were reduced by geldanamycin, an Hsp90 inhibitor. Mice inoculated with tumor cells stably transfected with mutant or wild-type luciferases were imaged. Increased light production and sensitivity were observed in the tumors bearing thermostable luciferase. Thermostable proteins increase imaging sensitivity. Presumably, as more active protein accumulates, detection is possible from a smaller number of mutant transfected cells compared to wild-type transfected cells. Mol Imaging (2004) 3, 324 – 332. Keywords: Luciferase, in vivo bioluminescence, thermostability, breast cancer tumors, metastasis.

Introduction In the current study, we test the hypothesis that thermostabilization will lead to a higher accumulation of firefly luciferase in human tumor cells and that this will improve the lower detection limit of luciferase expressing cells in whole animals. Luciferase is a powerful in vivo reporter system whose main strength is the relative absence of endogenous light-emitting signals from higher organisms. Hence, it can provide extremely high signal to noise. Over the past few years, methods have been developed to detect expression of luciferase proteins in living mice [1– 3]. Applications of this technology have grown, and now include the detection of tumor growth [4], the detection of tumor metastases [5], as a reporter for gene expression under control of regulable promoters [6 –9], as a reporter for efficacy of gene therapy [8,10 – 13], measurement of in vivo phar-

macodynamics [14,15] and as a reporter for toxicology studies [16 –18]. Luciferases are made by a wide variety of phyla, including bioluminescent species such as beetles, fireflies, bacteria, and marine coelenterates and dinoflagellates [19 –24]. The most widely used luciferase reporter gene is that of the North American firefly, Photinis pyralis. Wild-type, P. pyralis luciferase is thermolabile, with an in vitro half-life for activity on the order of 2– 3 min at 37C (see Results). However, luciferase is stabilized in vivo, with half-lives from 1 to 4 hr [25,26]. This has certain obvious advantages, as well as disadvantages. On the one hand, a high turnover rate of the reporter would be advantageous in time-sensitive studies, such as those to determine the response to an environmental toxin, infection, or induction or repression of gene expression. However, a high turnover rate could limit the accumulation of functional luciferase molecules and would undoubtedly have a higher explicit energy cost associated with continuous de novo protein synthesis. On the other hand, it is proposed that a lower turnover rate provided by thermostabilization will lead to higher accumulation of active luciferase molecules, and that this will lead to higher light output on a per-cell basis and thus improve detection. This would be advantageous for studies that are not time-sensitive, such as monitoring tumor growth or colonization of metastases. This is especially important in investigating metastases, as there would be significant interest in the fate of micro-metastases, which might otherwise escape detection due to their small size.

Materials and Methods Chemicals and Buffers Unless otherwise noted, all chemicals were obtained from Sigma (St. Louis, MO). Hanks Balanced Salt Solution Corresponding author: Robert J. Gillies, University of Arizona Cancer Center, Tucson, AZ 85724-5024; e-mail: [email protected]. *These authors contributed equally to this work. Received 10 November 2003; Accepted 17 August 2004. D 2005 Massachusetts Institute of Technology.

Thermostable Luciferases Baggett et al.

(HBSS): 5.4 mM KCl, 0.4 mM KH2PO4, 4.2 mM NaHCO3, 137 mM NaCl, 0.3 mM Na2HPO4. In order to generate HBSS at different pH values, pH was adjusted by adding either HCl or NaOH to the desired pH. Native firefly luciferase-recombinant protein was purchased from R&D Systems (Minneapolis, MN) (#700-LF-01M). Luciferin—D-luciferin, potassium salt, was a generous gift from Brian Ross (University of Ann Arbor, Michigan); Luciferase Assay reagent (LAR) and Cell Culture Lysis Reagent (CCLR) were purchased from Promega (Madison, WI). Luciferase mutants were cloned into a pET23a(+) bacterial expression vector at the NdeI and SalI sites of the MCS and maintained in Escherichia coli B121 DE3. Wild-type luciferase was isolated from pGEM-luc from Promega. The mammalian expression vector used was pcDNA3.1( ) from Invitrogen (Carlsbad, CA) with a selectable neomycin resistance gene. Production of Mutant Luciferases from Bacteria Preliminary experiments were done to optimize for induction time. DH5a bacteria were transformed with a pET23a(+)/mutant luciferase construct. Colonies of bacteria were isolated then grown in liquid cultures of LB broth (Sigma L 3152) in an orbital shaking incubator. One milliliter of cell suspension was removed, the bacterial cells pelleted, then resuspended in 1 mL HBSS, pH 7.8. This sample was placed in the luminometer, a background reading was taken, D-luciferin was added, and the light intensity was recorded. IPTG was then added to the remainder of the culture to a final concentration of 0.1 mM and cultures were left in an orbital shaking incubator. Samples were then removed at multiple time points post-IPTG induction, and the assay process was repeated. Peak induction was seen between 1 and 2 hr. In vitro experiments were performed by isolating the mutant luciferase protein from bacterial cultures. In an orbital shaking incubator, 2-mL starter cultures were grown from isolated colonies, incubated overnight, and then transferred to 100 mL LB broth. After incubating for 2 hr, cells were induced with IPTG. Two hours post-induction, cells were harvested and washed, a lysozyme buffer (9 mg/mL lysozyme in PBS) was added, and the cells were subjected to a freeze –thaw cycle after addition of Triton X-100. Samples were spun, and supernatant aliquoted, then frozen for use with in vitro temperature half-life assays.

325

(metastatic breast) and SW-480 (colon), were plated into six-well plates, or 10-cm dishes, for transfections. Stable cell lines were created by adding selection media (G418 sulfate, Geneticin, 400 mg/mL, Invitrogen), and cells maintained were under selection. Cells for mouse inoculations were expanded equally into several T-75 flasks, one flask was trypsinized and counted, in an appropriate number of flasks, cells were scraped, centrifuged, and resuspended in a volume of 1:1, saline/cell suspension to Matrigel (BD Biosciences, Franklin Lakes, NJ), so that a 100-mL inoculum contains 3 106 cells. Preparation of Mammalian Expression Vectors DNA was isolated from colonies of B121 DE3 E. coli transformed with a bacterial expression vector carrying one of six different mutant forms of luciferase that were previously generated by a random mutagenesis screen in yeast. These mutant sequences were cloned into the NdeI and SalI sites of the MCS of pET23a(+) (Tisi, unpublished). Competent DH5a cells were then transformed with the various mutants, designated A-F (see Table 1). Each mutant DNA was isolated, digested, and ligated into a mammalian expression vector. All mutants were transferred into the pcDNA3.1( ) mammalian expression vector at ApaI and XhoI. The wild-type luciferase was transferred from pGEM-luc (Promega) to pcDNA3.1( ), also at ApaI and XhoI. SW-480 colon cancer cells were transiently transfected using FuGene-6

Table 1. Thermostability of Luciferases Luciferase

Mutations

In vitro

In vivo +

T0.5 at 37C

DG (kJ/mol)

T0.5 at 37C

WT Mutant A

N/A E345K; T214A

3.06 8.5 ± 1.0

361.90 309.73

49.3 ± 21.4 136.3 ± 29.7

Mutant B

E345K; I232A; T214A

15.5 ± 1.45

353.40

141.9 ± 15.7

Mutant C Mutant D

E345K; A215L E345K; I232A;

7.36 ± 1.82 72.4 ± 3.1

403.65 270.74

160.4 ± 35.3 128.1 ± 11.4

82.1 ± 8.5

280.68

249.4 ± 58.7

75.1 ± 9.81

362.46

135.5 ± 16.9

T214A; F295L; S420T Mutant E

Mutant F

E345K; A215L; I232A; T214A; F295L E345K; A215L; I232A; T214A

For in vitro studies, luciferase mutants were cloned, expressed, and purified from DH5a bacteria and incubated at various temperatures (from 21C to 45C) for various times prior to assay at 23C. Plots of enzyme activity (light output) versus time were fitted to first-order exponential to obtain half-lives. Arrhenius plot of half-lives versus temperature was generated for each mutant to

Mammalian Cell Culture All mammalian cell lines were grown in DME/F-12 media (Sigma D0547) with 10% FBS. Cells were passaged weekly. The parental cell lines, MDA-mb-231 Molecular Imaging . Vol. 3, No. 4, October 2004

determine the DG+ associated with the inactivation process. For in vivo studies, SW-480 cells stably expressing each of these luciferases were treated with 0.01 mM cycloheximide at 37C and extracted at time points thereafter. The resulting plot of light output versus time was fitted to a first-order exponential to yield half-life.

326


(Roche, Indianapolis, IN) with each of the mammalian expression constructs. Transfected cells were incubated for 48 hr, lyzed with CCLR (Promega), and assayed for light. Polyclonal stable transfectants were also created by transfection of SW-480 cells or MDA-mb-231 cells using FuGene-6, 48-hr incubation, followed by selection using media containing 0.4 mg/mL G418. Cells continued to be passaged and grown in selection media. All stable lines were tested to confirm light production. Quantitative RT-PCR was conducted on RNA extracts of the various transfected cell lines to determine levels of message being produced by each cell line. Detection of Light Output in Cells and Extracts For most studies, light output was detected in vitro using an SLM 8100C spectrofluorometer operating without excitation. Photons were collected by Hamamatsu photomultipliers operating at room temperature, with attendant dark current. For all studies, the PMT voltage and gains were kept constant at 1250 V and 10, respectively, allowing comparison between assays run at different times. The reproducibility was verified in parallel experiments (data not shown). Detection of Light Output In Vivo Wild-type luciferase and thermostable mutants are to be compared to each other by imaging the cells in mice. Mutant luciferase cells were chosen for imaging studies based on mRNA expression levels being comparable to wild-type mRNA expression, as well as having a long half-life at 37C. Mutant F was chosen as the mutant luciferase for these studies. SCID mice were inoculated subcutaneously with MDA-mb-231/WT on the left flank and MDA-mb-231/mut F on the right flank using 3 106 cells in 100 mL Matrigel. On Day 8, tumors were barely palpable. The mice were anesthetized using a Ketamine, Acepromazine, and Xylazine cocktail (72 mg/ kg Ketamine, 6 mg/ kg each Acepromazine and Xylazine), then imaged using a VersArray 1300B cooled CCD camera (Roper Scientific, Tucson, AZ). A top-illuminated image was taken using a 100-msec exposure at f 16. The luminescent image was then taken using a 10-min exposure at f 2.8, 5 min after an intraperitoneal injection of 175 mg/ kg luciferin. An intensity map was created using WinView32 software (Princeton Instruments, Trenton, NJ). The final image was created by superimposing the color intensity map onto the light image using Adobe Photoshop 6.0. The imaging process was repeated on Days 15, 22, and 29, and caliper measurements taken of all tumors each time (tumors were only palpable, not measurable, until Day 22).

qRT-PCR Quantitative RT-PCR was done by isolating RNA from cell lines using a tRNA isolation mini-prep kit (Sigma, RTN-70), DNase treatment (Ambion, Austin, TX; Cat# 1906) of the samples, and followed by a SuperScript onestep RT-PCR reaction (Invitrogen). The RT-PCR reaction was run using a Smart Cycler (Cepheid, Sunnyvale, CA). Reactions were set up using previously designed PCR primers for luciferase (forward-GGGATACGACAAGGATATGGGC, reverse-TGGAACAACTTTACCGACCGC) and a GAPDH control, and detected by SYBR Green dye (Molecular Probes, Eugene, OR). Further experiments were carried out using a primer/probe design. Primers and a TET-labeled probe for luciferase (forward-GGCGCGTTATTTATCGGAGTT, reverse-TGGCGAGGGTGCTTACGT, probe-TET-TTGCGCCCGCGAACGACATT) were designed by using Primer Express software (Applied Biosystems, Foster City, CA). A commercially available FAM-labeled b-actin primer and probe set (Applied Biosystems) were used as an internal control. Data were expressed relative to glyceraldehydes phosphate dehydrogenase (GAPDH) or b-actin mRNA. Both of these have been shown to be robust control mRNA for these cell lines (Morse et al., submitted).

Results Determination of Optimal pH and Temperature of Wild-Type Luciferase The activity of native (wild-type) firefly luciferase is sensitive to both pH and temperature. Optimum pH was determined by incubating 20 mL of recombinant luciferase in 880 mL of HBSS adjusted to various pH levels. Baseline PMT current was monitored until stable, at which time 100 mL of LAR was added and the change in light output was recorded in arbitrary units. Light output was highest with the HBSS between pH 7 and 9. When these data were corrected for final pH, the peak range decreased significantly, as shown in Figure 1A. Consequently, HBSS at a pH of 7.8, prior to addition of LAR, was used for all subsequent studies, as the pH of the solution did not change upon addition of LAR. The effect of temperature on luciferase activity is illustrated in Figure 1B, which shows the time-dependent light output following addition of luciferase and LAR to HBSS and assayed at 23C and 40C. As shown in this figure, light output remains constant at 23C whereas less light is transiently emitted from the incubation at 40C. The thermosensitive component was determined by independently incubating LAR, HBSS, and luciferase at various temperatures prior to analysis. Neither LAR nor HBSS showed any thermolability (data not shown). Molecular Imaging . Vol. 3, No. 4, October 2004


327

Figure 1. Characterization of wild-type firefly luciferase in vitro and in vivo. (A) pH sensitivity. Commercial firefly luciferase was incubated at 23C in HBSS at various pH values and LAR induced light output was measured. Data are expressed as arbitrary units of bioluminescence ± SD (n 3 per datum). (B) Time course of light output at 23C and 40C. Luciferase and LAR were added to HBSS (pH 7.8) prewarmed to the indicated temperature and light captured by PMTs. Data are expressed as arbitrary units of bioluminescence. Gains, PMT voltages, and volumes were identical between samples. At 23C light was at a high intensity and remained stable. At 40C, intensity was not as high initially, and degraded over time. (C) Decrease in light output from as a function of time at different temperatures. Luciferase was incubated at the indicated temperatures for the indicated amount of time, after which it was cooled to 23C and assayed for light output by the addition of LAR. Data are expressed as arbitrary units of bioluminescence ± SD (n 3 per datum). (D) Half-lives at different temperatures with Arrhenius plot. The half-lives of luciferase were determined at different temperatures by exponential fit to data shown in C. An inverse plot (inset) indicates an energy of activation of approximately 340 kJ/mol. (E) Light output from cells following incubation at 27C and 37C. Equivalent cultures of 9L glioma cells were incubated for 18 hr at either 23C or 37C. Cells were trypsinized and suspended in a cuvette containing HBSS at 23C. For detection of in vivo luminescence, luciferin was added to the cell suspension and the luciferin induced light emission determined by PMTs. Alternatively, the in vitro luciferase activity was determined following lysis of the cells, and addition of LAR. Data are expressed as arbitrary units of bioluminescence ± SD (n = 4 per datum).

Molecular Imaging . Vol. 3, No. 4, October 2004

328


However, as shown in Figure 1C, luciferase itself was significantly thermolabile. Luciferase was incubated at different temperatures for various amounts of time in HBSS (pH 7.8), after which the temperature was reduced to 23C and the increase in light output was measured in response to addition of LAR. As shown in this figure, light output decreased with incubation time at all temperatures. The rates of decrease were multiphasic, suggesting complex mechanisms, yet were uniformly faster at higher temperatures. Luciferase activity was still present at up to 6 hr at 28C and 3 hr at 30C. At 34C and 37C, however, luciferase activity was completely abolished by 50 and 20 min, respectively. From these data, half-lives were calculated by fitting the data to simple first-order exponentials. The half-lives at different temperatures are shown in Figure 1D. An inverse plot of these data is shown in the inset. These data were fit to the Arrhenius equation, k = A exp(EA/RT), where k is the rate of enzyme inactivation, E is the activation energy associated with the inactivation process in kJ per mole, and T is the absolute temperature in Kelvin. This yields a DG‡ activation energy of 361.9 kJ/mol. An initial test was then conducted to assess whether the thermolability of luciferase significantly affected steady-state enzyme activity in vivo. Cultures of 9L glioma cells stably expressing WT luciferase were split and incubated in parallel for 18 hr at 23C or 37C with the expectation that lower temperatures would result in an accumulation of enzyme. At the end of the incubation, luciferase activity was assessed both in vivo with the addition of luciferin to whole cells and following cell lysis with the addition of LAR. As shown in Figure 1E, luciferase levels increased significantly upon incubation at 23C compared to 37C, supporting the hypothesis that luciferase thermolability significantly affects steady-state enzyme accumulation in vivo.

Thermostable Mutants The previous data suggest that the thermolability of luciferase could have a significant effect on steady-state protein levels in vivo. To further investigate this phenomenon, a series of thermostable mutants were analyzed (listed in Table 1). These were expressed and isolated from DH5a bacteria and the in vitro half-lives were determined as described in Materials and Methods. As shown in Table 1 and Figure 2B, the in vitro half-lives of these mutants were significantly longer as compared to wild-type luciferase. Note that the activation energy DG‡ of enzyme inactivation remained between 280 and 404 kJ/mol for all mutants, comparable to the 362 kJ/mol for wild type.

In Vivo Turnover Rates A more detailed analysis of the in vivo significance of thermolability was performed by determination of in vivo protein turnover rates. In these experiments, mutant luciferases were subcloned into mammalian expression vectors (pcDNA3.1( )) and stably expressed in SW-480 colon carcinoma cells under selection with G418. These cells were treated at 37C with 10 mM cycloheximide, which is sufficient to inhibit de novo protein synthesis. The amount of active luciferase was determined in vitro at different time points, and the subsequent rate of light reduction was used to estimate the in vivo turnover rate. A typical response is shown in Figure 2C, which shows the loss of light following cycloheximide treatment of SW-480 cells expressing luciferase mutant F. As shown in this figure, the light decreased with first-order exponential kinetics, with a calculated half-life of 135 min. Independent half-life determinations (with n = 1 sample per time point) were repeated at least four times per mutant and these data were averaged. A summary of results from all mutants is presented in Table 1. Note that the in vivo half-lives are significantly longer than those observed in vitro and that, in general, the relative in vivo half-lives compare favorably to the relative in vitro half-life values (Figure 2B). In other words, wild-type luciferase has the shortest half-life and mutant E has the longest half-life, both in vitro and in vivo. Western blot analyses show higher luciferase protein levels, therefore lower detection limits, in mutant luciferase compared to wild-type cell extracts (data not shown). Although the mechanisms behind the differences between in vivo and in vitro half-lives are not known, a reasonable hypothesis is that this is due to the presence of chaperone proteins in vivo [27,28]. This is supported by the observation that geldanamycin, an inhibitor of the chaperone, Hsp90, leads to a reduction in steady-state wild-type luciferase levels (Figure 2D). It should be noted that at this dose (200 nM), geldanamycin is not toxic to MDA-mb-231 cells (data not shown). Therefore, the decreased luciferase activity was not due simply to cell death.

In Vivo Imaging To test the hypothesis that thermostabilization of luciferases will lead to higher steady-state light output, mice were inoculated with tumor cells expressing either wild-type or mutant luciferases. In order for this comparison to be valid, the steady-state luciferase mRNA levels must be similar between two cell lines that are to be compared. SW-480 and MDA-mb-231 cells were transfected with mutant and wild-type luciferases, and mRNA Molecular Imaging . Vol. 3, No. 4, October 2004


329

Figure 2. Characterization of mutant luciferases in vitro and in vivo. (A) Renaturation of mutant luciferase protein. Thermostable mutant luciferase protein was incubated at 37C for various amounts of time. Light was either assayed immediately in the denatured state, or allowed to renature at room temperature prior to being assayed. (B) In vitro and in vivo half-lives of wild-type and mutant luciferases. In vitro half-lives were determined by IPTG induced bacteria and subsequent luciferase protein isolation. In vivo half-lives were determined from luciferase transfected mammalian tumor cell lines. (C) In vivo light output at times following cycloheximide. Typical result showing effects of cyclohexamide inhibiting protein synthesis in cells transfected with thermostable mutant luciferase. (D) Effects of geldanamycin. This suggests the role of chaperone activity in the stabilization of luciferase protein. Geldanamycin inhibits chaperone activity, thereby decreasing light output, without being toxic to the cell.

levels were compared by quantitative RT-PCR. In initial experiments, primers for GAPDH were used to compare luciferase levels by taking the log2 amount of luciferase mRNA expressed relative to the log2 GAPDH product from the same preparation. Subsequent experiments used a primer/probe design to assess, by multiplex analysis, b-actin and luciferase levels within a single reaction. Steady-state luciferase mRNA levels were compared across all cell lines and there was a general agreement between mRNA and light output (data not shown). Of all cell lines, SW-480 cells expressing mutant E and MDA-mb-231 cells expressing mutant F had mRNA levels that were closest to their respective control cell lines expressing wild-type luciferases (data not shown). Although both of these were tested in vivo, subsequent RT-PCR from tumor extracts showed that SW-480 cells expressing wild-type luciferase consistently lost mRNA Molecular Imaging . Vol. 3, No. 4, October 2004

expression when grown in the absence of selection (data not shown). Consequently, in vivo studies were carried out using MDA-mb-231 transfected with WT and mutant F luciferases, which were shown to have comparable mRNA levels and representative differences in their steady-state light output, as seen in Figure 3A. Transfected WT or mutant F MDA-mb-231 cells were grown as subcutaneous xenografts in flanks of SCID mice imaged during tumor growth. All mutant tumors showed greater light intensity than wild-type tumors. SimplePCI software was used to analyze the light intensity from bioluminescent images. Images were converted to 16-bit TIF images, and were analyzed by implementing a threshold of twice background to automatically select regions of light output. The selected regions were measured for number of pixels in a region, total light intensity over the region,

330


Figure 3. (A) qRT-PCR. Using a primer/probe design, luciferase levels were compared to b -actin levels. Results of MDA-mb-231 cells transfected with either wild-type or mutant F luciferase that were used to inoculate mice are shown here. Samples are compared at a cycle threshold (30 units), the point at which the slope of the curve changes. At this point the ratio of luciferase to b -actin shows that the expression of luciferase in WT is slightly higher than that of the mutant, yet not significantly different. (B) Light output of WT and mutant F from tumors in vivo. SimplePCI software was used to analyze images. A threshold was set to automatically select regions of increased intensity. The areas selected were calculated for number of pixels in the region and total intensity within the region. Area in pixels is an approximation of tumor volume, since tumors were not large enough for caliper measurements until Week 3. (C) In vivo imaging sensitivity. Tumor volumes were initially measured by MR imaging before they were large enough for caliper measurements. Tumor volumes here were estimated based on MRI tumor volumes versus number of detectable pixels for the remainder of the experiment; only a percentage of the wild-type tumor is detectable compared to the detectable area of the mutant tumor. Due to increased sensitivity, mutant luciferase can be imaged at a much smaller size, with much greater intensity. Tumor volumes through 4 weeks postinoculation showed that wild-type tumors were larger than mutant F tumors. (D) Bioluminescence of mouse bearing tumors expressing WT and mutant luciferases. In all cases, the mutant luciferase showed greater light output than WT. In this image, MR images were taken to determine tumor volumes, showing that although the WT tumor is four times larger than the mutant tumor, the mutant tumor has a higher light intensity, greater light per pixel, and greater detectable area.

and an average of intensity per pixel was calculated (Figure 3B). As shown in Figure 3D, the light output from the tumor-expressing mutant F is significantly higher than that of the tumor-expressing wild-type luciferase. In all mice imaged, the mutant tumor produced greater light intensity than that of the wild-type tumor on all days imaged. Due to slightly varying tumor

growth rates, most animals had larger WT tumors than mutant tumors over the imaging time course, whereas a few had larger mutant F tumors than WT tumors. To verify that tumor size was not the influencing factor, MR images were taken to calculate the tumor volume, before the tumors were palpable. This showed that tumor volume was not the reason for increased light intensity. Molecular Imaging . Vol. 3, No. 4, October 2004


The MR tumor volume measurements were also useful, since imaging of tumors was possible more than 2 weeks before caliper measurements could be made.

Discussion Luciferase is a powerful reporter gene for in vitro detection of gene expression and has more recently been used in in vivo bioluminescence imaging. Because the wild-type protein is relatively thermolabile at in vivo temperatures, it was hypothesized that this would limit the in vivo accumulation of light emitting activity on a per-cell basis. This work has shown that thermostabilization achieved through specific mutations and can lead to higher steady-state levels of enzyme accumulation in vivo. In all cases, heat-induced loss of enzyme activity was associated with an energy of activation near 340 kJ/ mol. The relationship between the measured in vivo half-lives and the apparent activation energy of inactivation is clearly complex. For example, wild type and mutant F have, essentially, the same value of DG‡ (361.9 and 362.5 kJ/mol, respectively), yet the in vivo half-life of mutant F is close to three times that of WT. This may reflect that the irreversible inactivation of luciferases can occur via several pathways [29 – 32], making interpretations of apparent DG‡ values difficult. These analyses indicated largely similar activation energies for enzyme inactivation, for both mutants and wildtype enzymes, in the range of 340 kJ/mol. The fact that the mutant luciferases show greatly increased half-lives despite little variation in their activation energies for inactivation is noteworthy. This suggests that the mechanisms by which the various mutations stabilize the enzyme do not simply come from enthalpically favorable interactions increasing the activation energy for inactivation. Although all mutants express longer halflives compared to wild-type luciferase both in vitro and in vivo, the in vivo half-lives of all enzymes were significantly longer than those observed in vitro. The mechanisms underlying the enhanced in vivo thermostability are not known. One factor is likely the activity of chaperone proteins, which could act to inhibit the transition from native to molten globule state [33]. Indeed, it has previously been shown that chaperones prolong the half-lives of luciferases both in cell-free and cell-based assays [27 – 29,33]. Consistent with this, we have shown that geldanamycin, an inhibitor of the chaperone, Hsp90, reduces steady-state luciferase activities in vivo. Notably, the in vivo stability can also be enhanced pharmacologically with phenylbenzothiazole, a competitive inhibitor of luciferin binding [34]. Hence, in vivo stability might be coupled to activity, yet this is Molecular Imaging . Vol. 3, No. 4, October 2004

331

not explicitly the case as the specific activities of the thermostable luciferases are comparable to wild type. In all cases, thermostabilized luciferases accumulated to higher levels than wild type, as determined by the relationship between luciferin-induced light output (activity) and mRNA levels. Notably, these data do not address the effects of these mutations on the reaction kinetics, which are being investigated elsewhere. Nonetheless, the simplest explanation for the observed results is the increased accumulation of luciferase enzyme with increased half-lives at in vivo temperatures. In either case, the empirically observed increase in light output by the cells and tumors transfected with thermostabilized luciferase increased its efficiency as an in vivo reporter gene. Fewer numbers of cells can be imaged, therefore greatly increasing sensitivity. Previously, the detection limit for bioluminescence imaging was determined to be approximately 1000 cells [1,26]. The in vivo bioluminescent data here suggest that the minimum detectable cell number would be closer to 100 cells. Hence, the use of thermostabilized luciferases may allow monitoring of micro-metastases and the early stages of tumor growth.

Acknowledgments We thank Prof. Thomas Baldwin for helpful discussions in the preparation of this manuscript, and Ms. Merry Warner for help in the production of this manuscript. The study was supported by the Hughes and NSF-supported Undergraduate Biology Research Program (RR, SM, SM), and NIH CA77375 (BB, RJG).

References [1] Contag CH, Jenkins D, Contag PR, Negrin RS (2000). Use of reporter genes for optical measurements of neoplastic disease in vivo. Neoplasia. 2:41 – 52. [2] Contag CH, Spilman SD, Contag PR, Oshiro M, Eames B, Dennery P, Stevenson DK, Benaron DA (1997). Visualizing gene expression in living mammals using a bioluminescent reporter. Photochem Photobiol. 66:523 – 531. [3] Contag PR, Olomu IN, Stevenson DK, Contag CH (1998). Bioluminescent indicators in living mammals. Nat Med. 4: 245 – 247. [4] Rehemtulla A, Stegman LD, Cardozo SJ, Gupta S, Hall DE, Contag CH, Ross BD (2000). Rapid and quantitative assessment of cancer treatment response using in vivo bioluminescence imaging. Neoplasia. 2:491 – 495. [5] Wetterwald A, van der PG, Que I, Sijmons B, Buijs J, Karperien M, Lowik CW, Gautschi E, Thalmann GN, Cecchini MG (2002). Optical imaging of cancer metastasis to bone marrow: A mouse model of minimal residual disease. Am J Pathol. 160:1143 – 1153. [6] Shibata T, Giaccia AJ, Brown JM (2000). Development of a hypoxia-responsive vector for tumor-specific gene therapy. Gene Ther. 7:493 – 498. [7] Shibata T, Giaccia AJ, Brown JM (2002). Hypoxia-inducible

332

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20] [21]

Thermostable Luciferases Baggett et al. regulation of a prodrug-activating enzyme for tumor-specific gene therapy. Neoplasia. 4:40 – 48. Shin JH, Yi JK, Lee YJ, Kim AL, Park MA, Kim SH, Lee H, Kim CG (2003). Development of artificial chimerical gene regulatory elements specific for cancer gene therapy. Oncol Rep. 10: 2063 – 2069. Chinnusamy V, Stevenson B, Lee BH, Zhu JK (2002). Screening for gene regulation mutants by bioluminescence imaging. Sci STKE. 2002:L10. Rehemtulla A, Hall DE, Stegman LD, Prasad U, Chen G, Bhojani MS, Chenevert TL, Ross BD (2002). Molecular imaging of gene expression and efficacy following adenoviral-mediated brain tumor gene therapy. Mol Imaging. 1:43 – 55. Chaudhuri TR, Rogers BE, Buchsbaum DJ, Mountz JM, Zinn KR (2001). A noninvasive reporter system to image adenoviralmediated gene transfer to ovarian cancer xenografts. Gynecol Oncol. 83:432 – 438. Brun S, Faucon-Biguet N, Mallet J (2003). Optimization of transgene expression at the posttranscriptional level in neural cells: Implications for gene therapy. Mol Ther. 7:782 – 789. Siemens DR, Crist S, Austin JC, Tartaglia J, Ratliff TL (2003). Comparison of viral vectors: Gene transfer efficiency and tissue specificity in a bladder cancer model. J Urol. 170:979 – 984. Nunn C, Feuerbach D, Lin X, Peter R, Hoyer D (2002). Pharmacological characterisation of the goldfish somatostatin sst5 receptor. Eur J Pharmacol. 436:173 – 186. Kimm-Brinson KL, Moeller PD, Barbier M, Glasgow H, Jr, Burkholder JM, Ramsdell JS (2001). Identification of a P2X7 receptor in GH(4)C(1) rat pituitary cells: A potential target for a bioactive substance produced by Pfiesteria piscicida. Environ Health Perspect. 109:457 – 462. Catania JM, Parrish AR, Kirkpatrick DS, Chitkara M, Bowden GT, Henderson CJ, Wolf CR, Clark AJ, Brendel K, Fisher RL, Gandolfi AJ (2003). Precision-cut tissue slices from transgenic mice as an in vitro toxicology system. Toxicol In Vitro. 17:201 – 205. Billard P, DuBow MS (1998). Bioluminescence-based assays for detection and characterization of bacteria and chemicals in clinical laboratories. Clin Biochem. 31:1 – 14. Bitton G, Koopman B (1992). Bacterial and enzymatic bioassays for toxicity testing in the environment. Rev Environ Contam Toxicol. 125:1 – 22. Greer LF, Szalay AA, III (2002). Imaging of light emission from the expression of luciferases in living cells and organisms: A review. Luminescence. 17:43 – 74. Hastings JW (1996). Chemistries and colors of bioluminescent reactions: A review. Gene. 173:5 – 11. Viviani VR (2002). The origin, diversity, and structure func-

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

tion relationships of insect luciferases. Cell Mol Life Sci. 59: 1833 – 1850. Lall AB, Ventura DS, Bechara EJ, de Souza JM, Colepicolo-Neto P, Viviani VR (2000). Spectral correspondence between visual spectral sensitivity and bioluminescence emission spectra in the click beetle Pyrophorus punctatissimus (Coleoptera: Elateridae). J Insect Physiol. 46:1137 – 1141. Viviani VR, Silva AC, Perez GL, Santelli RV, Bechara EJ, Reinach FC (1999). Cloning and molecular characterization of the cDNA for the Brazilian larval click-beetle Pyrearinus termitilluminans luciferase. Photochem Photobiol. 70:254 – 260. Schmitter RE, Njus D, Sulzman FM, Gooch VD, Hastings JW (1976). Dinoflagellate bioluminescence: A comparative study of in vitro components. J Cell Physiol. 87:123 – 134. Day RN, Kawecki M, Berry D (1998). Dual-function reporter protein for analysis of gene expression in living cells. BioTechniques. 25:848 – 4, 856. Leclerc GM, Boockfor FR, Faught WJ, Frawley LS (2000). Development of a destabilized firefly luciferase enzyme for measurement of gene expression. BioTechniques. 29:590 – 596, 598. Souren JE, Wiegant FA, van Hof P, van Aken JM, van Wijk R (1999). The effect of temperature and protein synthesis on the renaturation of firefly luciferase in intact H9c2 cells. Cell Mol Life Sci. 55:1473 – 1481. Souren JE, Wiegant FA, van Wijk R (1999). The role of hsp70 in protection and repair of luciferase activity in vivo; experimental data and mathematical modelling. Cell Mol Life Sci. 55: 799 – 811. Herbst R, Schafer U, Seckler R (1997). Equilibrium intermediates in the reversible unfolding of firefly (Photinus pyralis) luciferase. J Biol Chem. 272:7099 – 7105. Herbst R, Gast K, Seckler R (1998). Folding of firefly (Photinus pyralis) luciferase: Aggregation and reactivation of unfolding intermediates. Biochemistry. 37:6586 – 6597. Baldwin TO, Chen LH, Chlumsky LJ, Devine JH, Ziegler MM (1989). Site-directed mutagenesis of bacterial luciferase: Analysis of the ‘‘essential’’ thiol. J Biolumin Chemilumin. 4:40 – 48. Chen LH, Baldwin TO (1989). Random and site-directed mutagenesis of bacterial luciferase: Investigation of the aldehyde binding site. Biochemistry. 28:2684 – 2689. Flynn GC, Beckers CJ, Baase WA, Dahlquist FW (1993). Individual subunits of bacterial luciferase are molten globules and interact with molecular chaperones. Proc Natl Acad Sci U.S.A. 90:10826 – 10830. Thompson JF, Hayes LS, Lloyd DB (1991). Modulation of firefly luciferase stability and impact on studies of gene regulation. Gene. 103:171 – 177.

Molecular Imaging . Vol. 3, No. 4, October 2004