Haemoglobin interferes with the ex vivo luciferase luminescence assay: consequence for detection of luciferase reporter gene expression in vivo

Abstract

The luciferase reporter gene is a useful tool for determining the efficacy of transfection of plasmid DNA and adenovirus-mediated gene transfer in vivo. However, we report here that the haemoglobin present in tissue samples can mask the detection of the luciferase activity and lead to underestimation of the luciferase levels. We evaluated the degree of interference in different organ samples of mice and investigated the possibilities for removal of haemoglobin from tissue samples by: (1) perfusion of the whole animal; (2) different hypotonic treatments lysing preferentially red blood cells; and (3) chromatographic separation. Removal of haemoglobin resulted in significantly improved detection of luciferase activity from tissue samples. The results indicate that the luciferase activity determined in tissue samples may not reflect the actual level of reporter gene expression, if contaminating blood is not taken into consideration.

Main

Luminescent reporter gene assays for luciferase are frequently employed to determine the efficiency of gene transfer in vitro, as they allow quantitative evaluation of luciferase gene expression after transfection. Numerous studies in vivo have used this assay to analyse the gene transfer efficacy of plasmid DNA by different liposome preparations in vivo1234567 and some investigations have also examined the efficacy of transduction by adenoviral vectors carrying the luciferase gene after administration to mice by different routes.89.

After performing successful nonviral integrin-mediated gene transfer by an RGD-oligolysine vector using a luciferase reporter gene in vitro101112 we started testing this vector system in vivo in mice. The extremely low levels of luciferase activity in these experiments and similar reports in the literature123456 led us to question the sensitivity of detection of luciferase activity by the assay generally employed for animal tissues and to investigate possible factors that may interfere with the luciferase assay, in particular whole blood.

In initial model experiments we added different amounts of whole mouse blood, haemoglobin or plasma to a constant amount of lysed tracheal cells expressing a luciferase reporter gene, and evaluated the luciferase activity by luminometry. Both whole blood and haemoglobin, the latter at a concentration approximately equivalent to that of whole mouse blood led to a similar decrease in the detected luminescence (Figure 1). Neither plasma (blood collected in the presence of EDTA thus containing coagulation factors) (Figure 1) nor serum (clotted blood devoid of coagulation factors) (data not shown), added in the same volumes as whole blood, resulted in a significant decrease in the relative light units (RLU). Addition of lysate buffer instead of blood or haemoglobin did not modify the control value (data not shown). These results indicate that it is the haemoglobin in the blood which interferes with the assay and suggest that haemoglobin contamination results in a considerable decrease in the detectable luciferase activity.

Figure 1
figure1

Blood components interfere with the luciferase assay. Different amounts of whole mouse blood (□), haemoglobin (140 mg/ml) (▪) or plasma () were added to a constant amount of cell lysate (50 μl) following transfection with a luciferase reporter gene plasmid and the relative light units (RLU) were determined using a luminometer. Fetal human tracheal epithelial cells, 56FHTE8o, were transfected with a vector complex composed of the American firefly (Phoptinus pyralis) luciferase gene plasmid pGL3 (Promega, Charbonnières, France), a [K]16RGD peptide and the cationic lipid lipofectamine at a ratio of 1:5:18 (w:w:w), respectively. Cells were harvested in lysis buffer (25 mM Tris/H3PO4, pH 7.8, 1% Triton X-100, 15% glycerol, 1 mM EDTA, 10 mM MgCl2, 1 mM DTT) for 15 min at 4°C. The cell lysate was then transferred to 1.5 ml Eppendorf tubes and centrifuged for 10 min at 21700 g. Luciferase activity of the cell lysate was then determined on aliquots (100 μl) using a Berthold luminometer (Lumat LB9507, Evry, France) with a 10 s integration period after automatic injection of 100 μl of a D-luciferin solution (lysis buffer without DTT, with 100 mM ATP, 43 mg/ml D-luciferin). The spectral sensitivity of the photomultiplier covers a range between 390 and 620 nm. Results are representative of three individual experiments each performed in triplicate.

Theoretically it may be possible to calculate the degree of interference if the haemoglobin content of the different organs is known. We therefore estimated the haemoglobin content of organ samples by colorimetry using Drabkin's reagent. The degree of interference could then be determined based on the data shown in Figure 1. It was highest for the lung and slightly lower for both the liver and the spleen (Figure 2). Chloroform extraction before analysis was required to remove lipid which interfered with the colorimetric estimations. During this procedure some loss of haemoglobin occurred, which may have resulted in a slight underestimation of the amount of haemoglobin.

Figure 2
figure2

Percentage of interference with the luciferase assay by haemoglobin in homogenates of mouse organs. The theoretically possible interference was calculated from the data in Figure 1 and from the estimation of the haemoglobin content estimated colorimetrically using Drabkin's reagent of a kit for total haemoglobin quantification (Sigma Aldrich, St Quentin Fallavier, France). The percentage decrease in RLU/mg was plotted against the amount of haemoglobin and the percentage interference determined knowing the haemoglobin content of the different samples. Liver, lung and spleen were homogenised in lysis buffer (as above) at 4°C and centrifuged for 10 min at 21700 g. The supernatant was recovered, lipid extraction with chloroform:sample (1:1, v:v) performed and an aliquot used for haemoglobin estimation.

Having established that haemoglobin interferes with the assay, we hypothesised that dilution of the samples may reveal a greater amount of luciferase activity. Indeed, a dilution factor of 10 for the liver and spleen homogenates led to a seven- and 2.8-fold increase in detectable luciferase activity, respectively, and a dilution factor of four for the lung led to a 1.7-fold increase (Figure 3).

Figure 3
figure3

Dilution of organ homogenates results in an increase in the detectable luciferase activity. Mice were injected i.v. (tail vein) with 108 decp50 of a type 5 adenovirus obtained from Trangene SA, Strasbourg, France, carrying a luciferase gene (AdMLPLuc) under the major late promoter and killed after 72 h with a lethal dose of ketamine. Liver, spleen and lungs were removed, rinsed in PBS, diced and half of each organ was homogenised in 1 ml lysis buffer. The luciferase activities of aliquots (100 μl) of liver, spleen and lung homogenates undiluted or diluted by a factor of 2, 4 and 10 were determined as described in Figure 1. Results are the mean ± s.e.m. for homogenates from five mice. Ketamine at concentrations estimated to be present in organs of the mice does not affect the luciferase activity (data not shown).

Although the emission spectrum of firefly luciferase peaks at 562 nm (SWISS-PROT entry LUCI PHOPY), it ranges from 510 to 650.13 The major peak of haemoglobin light absorption occures at 408 nm, as indicated above for haemoglobin quantification. However, haemoglobin exists in several different forms (O2Hb, COHb, HHb and MetHb) each of which has different absorbance spectra. We examined the light absorbance spectra of the different mice organ homogenates, haemoglobin and mice whole blood to determine the degree of overlap with the emission spectrum of luciferase (Figure 4). Substantial absorbance at the peak values of 575 and 540 nm was observed in diluted homogenates. Thus these forms of haemoglobin are probably responsible for the interference since they overlap with the luciferase spectrum.

Figure 4
figure4

Visible absorption spectra of mouse organ homogenates. The spectra of liver (a) (1/375 dilution), lung (b) (1/125 dilution) and spleen (c) (1/125 dilution) homogenates and of mouse haemoglobin (d) (1.4 mg/ml) and mouse whole blood (e) (1/75 dilution) are presented. The bar represents the range of the luciferase emission spectrum as described previously.13

Next therefore, we sought to remove haemoglobin from the tissue samples before determination of the luciferase activity. Three different approaches were investigated: (1) transcardiac perfusion of mice with an isotonic buffer; (2) preferential lysis of red blood cells with hypotonic buffers; and (3) chromatographic separation of the luciferase protein from haemoglobin in homogenates.

Transcardiac perfusion was performed through the left and right ventricles and resulted in visible removal of blood from the liver and lungs. This is reflected in the decrease in the estimated haemoglobin content of these organs (Figure 5a and b). Measurement of the optical density at 408 nm, the optimum wavelength for haemoglobin light absorption, is an easy and rapid method for determining the efficacity of the perfusion. The lung was the most efficiently perfused organ. This procedure also resulted in a substantial increase in the detectable luciferase activities, in particular for the liver (six-fold) and the lung (eight-fold) (Figure 5c) while the spleen showed a lower increase (two-fold).

Figure 5
figure5

Transcardiac perfusion results in decreased haemoglobin content and increased detectable luciferase activity in mouse organ homogenates. Mice were injected into the tail vein with 100 μl (5 × 1010 p.f.u.) of AdMLPLuc, 48 h later anaesthetised terminally and perfused with 2 × 20 ml of PBS containing 2000 units of heparin/500 ml over a period of about 5 min, first through the left ventricle and then through the right ventricle. Organs were homogenised in lysis buffer. (a) Haemoglobin content estimated by spectroscopy at 408 nm. (b) Haemoglobin content estimated by colorimetry using Drabkin's reagent. (c) Luciferase activity determined as described in Figure 1. Control (full bars), perfused (hatched bars). The protein concentration of the homogenates was determined using the Bradford method.16 Results are the mean ± s.e.m. for six mice. Statistical analysis was performed using the nonparametric Mann–Whitney U test. *Probability values <0.05 were considered statistically significant.

We then investigated two different hypotonic treatments allowing preferential lysis of red blood cells for removal of haemoglobin. Such techniques are used to remove erythrocytes from lymphocyte and spleen cell preparations.14

To achieve this, tissues of mice injected with AdMLPLuc were diced very finely and incubated on ice for a short period in either: (1) PBS; or (2) a hypotonic Tris-HCl buffer containing either 5 mM MgCl2 and 10 mM NaCl or 144 mM NH4Cl. The supernatant containing haemoglobin was removed and the remaining pellet was washed twice before lysis and assay for luciferase activity. This hypotonic treatment resulted in a slight increase in the detected luciferase activity in the liver (1.3-fold), but also in reduced activity in the spleen and lung samples which may be due to the loss of some luciferase activity during hypotonic treatment.

The separation of the luciferase protein from haemoglobin was also attempted by other techniques including ammonium sulphate precipitation and chromatography. Firefly luciferase has been reported to salt-out at 50 to 60% (NH4)2SO4,13 while haemoglobin precipitates at 85%.15 However, when an aliquot of tissue homogenates in lysis buffer was brought to 60% (NH4)2SO4 the precipitate still contained some haemoglobin, and in the case of the liver it was no longer soluble in lysis buffer. As firefly luciferase has a lower pI than haemoglobin ion-exchange chromatography was also tested but, unfortunately, substantial separation was not achieved. In addition, high salt concentrations (200 mM MgCl2) were required to elute the luciferase from the anion exchanger and we noted that the luminescence reaction was sensitive to the salt concentration. Since haemoglobin contains twice as many histidine residues as luciferase we also tested immobilized metal affinity resines usually used for purification of histidine-tagged proteins. However, we were unable to separate the two proteins satisfactorily even by elution with imidazol which competes with histidine for binding. Thus all these alternative techniques investigated appear less efficient than transcardiac perfusion.

The observed decrease in the detectable luciferase activity in the presence of haemoglobin is most likely due to quenching of the luminescence that is emitted by the reaction of luciferase with its substrate D-luciferin, rather than to direct inhibition of the luciferase activity by haemoglobin. This is suggested by the increase in the detectable luciferase activity after dilution of organ homogenates which decreases the absorption at wavelengths corresponding to the emission peak of luciferase (562 nm; yellow–green). However, direct interaction between the two proteins can not be excluded.

There exists several luciferases other than the North American firefly luciferase used in the present study. Each has a different emission maxima: the North American firefly, Photinus pyralis (LUCI PHOPY); Japanese firefly, Luciola cruciata (LUCI LUCCR); the sea pansy, Renilia reniformis (LUCI RENRE); the sea firefly, Vargula hilgendorfii (LUCI VARHI) have lambda max at 562, 544, 480 and 460 nm, respectively. Detection of the activity of all these enzymes would be reduced in the presence of haemoglobin. However, mutant Japanese firefly luciferase producing orange (lambda max, 607 nm) and red light (lambda max, 609 and 612 nm) have been reported and thus their emission should not be quenched by haemoglobin,17 though the detector used would also be required to detect in the red. The spectral sensitivity of the photomultiplier of the luminometer used in this study covers a range between 390 and 620 nm with greatest sensitivity in the blue range. This instrumentation is also the most frequently used by most investigators interested in evaluating gene transfer. In addition, other broad spectrum detectors used in low light detection such as charge coupled devices (CCD) cameras may compensate for some of the effect of hemoglobin on yellow–green luciferase detection.

Nonetheless our results indicate that it is advisable to remove a maximum amount of haemoglobin when performing quantitative analysis of luciferase activity in organs. The perfusion of the whole animal appears to be an effective procedure for removal of haemoglobin from organs. We were particularly interested in the lung and therefore chose to perform intracardiac perfusion. This has allowed us to establish the principle of perfusion as a technique of circumventing the interference of haemogolobin in the luciferase assay. However, perfusion of the liver is probably best performed directly through the portal vein.

In conclusion, our results suggest that luciferase activity levels in previous reports in vivo may have been underestimated due to the presence of blood.

References

  1. 1

    Thierry AR et al. Systemic gene therapy: biodistribution and long-term expression of a transgene in mice Proc Natl Acad Sci USA 1995 92: 9742–9746

    CAS  Article  Google Scholar 

  2. 2

    Fortunati E et al. In vitro and in vivo gene transfer to pulmonary cells mediated by cationic liposomes Biochim Biophys Acta 1996 1306: 55–62

    Article  Google Scholar 

  3. 3

    Hara T, Tan Y, Huang L . In vivo gene delivery to the liver using reconstituted chylomicron remnants as a novel nonviral vector Proc Natl Acad Sci USA 1997 94: 14547–14552

    CAS  Article  Google Scholar 

  4. 4

    Li S, Huang L . In vivo gene transfer via intravenous administration of cationic lipid-protamine-DNA (LPD) complexes Gene Therapy 1997 4: 891–900

    CAS  Article  Google Scholar 

  5. 5

    Liu F, Qi H, Huang L, Lui D . Factors controlling the efficiency of cationic lipid-mediated transfection in vivo via intravenous administration Gene Therapy 1997 4: 517–523

    CAS  Article  Google Scholar 

  6. 6

    Barron LG et al. Effects of complement depletion on the pharmacokinetics and gene delivery mediated by cationic lipid-DNA complexes Hum Gene Ther 1998 9: 315–323

    CAS  Article  Google Scholar 

  7. 7

    Goula D et al. Polyethylenimine-based intravenous delivery of transgenes to mouse lung Gene Therapy 1998 5: 1291–1295

    CAS  Article  Google Scholar 

  8. 8

    Huard J et al. The route of administration is a major determinant of the transduction efficiency of rat tissues by adenoviral recombinants Gene Therapy 1995 2: 107–115

    CAS  Google Scholar 

  9. 9

    Yei S et al. Adenovirus gene transfer for cystic fibrosis: quantitative evaluation of repeated in vivo vector administration to the lung Gene Therapy 1994 1: 192–200

    CAS  PubMed  PubMed Central  Google Scholar 

  10. 10

    Colin M et al. Liposomes enhance delivery and expression of an RGD-oligolysine gene transfer vector in human tracheal cells Gene Therapy 1998 5: 1488–1498

    CAS  Article  Google Scholar 

  11. 11

    Colin M et al. Cell delivery, intracellular trafficking and expression of an integrin-mediated gene transfer vector in cultured tracheal epithelial cells Gene Therapy 2000 7: 139–152

    CAS  Article  Google Scholar 

  12. 12

    Harbottle RP et al. An Arg-Gly-Asp/oligolysine peptide: a prototype construct for integrin mediated gene delivery Hum Gene Ther 1998 9: 1037–1047

    CAS  Article  Google Scholar 

  13. 13

    Wood KV, Lam YA, McElroy WD . Introduction to beetle luciferase and their applications J Biolum Chemilum 1989 4: 289–301

    CAS  Article  Google Scholar 

  14. 14

    Leenen PJ et al. Heterogeneity of mouse spleen dendritic cells: in vivo phagocytic activity, expression of macrophage markers, and subpopulation turnover J Immunol 1998 160: 2166–2173

    CAS  Google Scholar 

  15. 15

    Scher W, Holland JG, Friend C . Hemoglobin synthesis in murine virus-induced leukemic cells in vitro. I. Partial purification and identification of hemoglobins Blood 1971 37: 428–437

    CAS  PubMed  Google Scholar 

  16. 16

    Bradford M . A rapid and sensitive method for the quantification of microgram quantities of protein utilizing protein-dye binding Anal Biochem 1976 72: 248–254

    CAS  Article  Google Scholar 

  17. 17

    Kajiyama N, Nakano E . Isolation and characterization of mutants of firefly luciferase which produce different colors of light Protein Eng 1991 6: 691–693

    Article  Google Scholar 

Download references

Acknowledgements

This work was supported by grants from the ‘Institut National de la Santé et de la Recherche Médicale’ (INSERM) and from the ‘Association Française de Lutte contre la Mucoviscidose’. We thank Dr A Pavirani of Transgene SA (Strasbourg, France) and Dr M Themis (Imperial College, London) for providing AdMLPLuc and Dr D Gruenert (Gene Therapy Core Center, USA) for the 56FHTE8o cell line. CC is funded by the Muller Bequest/CF-Trust, the MRC and the March of Dimes Birth Defects Foundation. HS was supported by the Deutsche Forschungsgemeinschaft.

Author information

Affiliations

Authors

Corresponding author

Correspondence to MC Brahimi-Horn.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Colin, M., Moritz, S., Schneider, H. et al. Haemoglobin interferes with the ex vivo luciferase luminescence assay: consequence for detection of luciferase reporter gene expression in vivo. Gene Ther 7, 1333–1336 (2000). https://doi.org/10.1038/sj.gt.3301248

Download citation

Keywords

  • adenovirus
  • gene therapy
  • haemoglobin
  • luciferase
  • perfusion
  • red blood cells

Further reading

Search

Quick links