Immune-mediated destruction of transfected myocytes following DNA vaccination occurs via multiple mechanisms

Abstract

The delivery of antigenic proteins in the context of a DNA vaccine leads to the intracellular synthesis of antigen and the induction of both humoral and cellular immune responses. Subsequent to immune activation, any transfected cell expressing the immunogenic protein should, by the rules of immunology, become a legitimate target for removal by immune-mediated mechanisms. Herein, we have used an indirect assay of myocyte integrity following intra-muscular (i.m.) delivery of a DNA vaccine, in mice with various immune deficiencies, to determine which immunological mechanisms may be involved in destruction of antigen-expressing cells. We demonstrate that destruction of antigen-expressing myocytes following i.m. injection of a DNA vaccine is dependent on major histocompatability complex (MHC) class II restricted CD4+ T cell activation, but is not mediated solely by MHC I-restricted or perforin-mediated lysis and appears to have a component that is antibody-mediated. Although we studied myocytes, the results likely represent what happens to any transfected cell expressing a foreign antigen. This study underscores the ability of DNA vaccines at inducing antigen-specific immune responses that include a number of effector mechanisms. From the perspective of gene therapy, this study highlights the significance of immune activation when considering strategies where maintenance of therapeutic gene expression is desired. Gene Therapy (2001) 8, 1395–1400.

Introduction

We have developed a DNA vaccine against the hepatitis B virus (HBV) based on expression of HBsAg.1,2 I.M. delivery of this vaccine in mice induces strong and long-lasting immune responses comprising both antigen-specific antibodies (Abs) and cytotoxic T lymphocytes (CTL),1,3 similar to those seen in various other models.4

The finding that a single administration of a DNA vaccine to mice results in life-long immunity5 was initially thought to be due to the prolonged expression of antigen, as had been observed previously with certain reporter genes such as luciferase.6,7 However, in the presence of a developing immune response, characterized by the generation of strong MHC class I restricted CTL,3 it would seem likely that transfected myocytes and other cells that express antigen would become potential targets for CTL-mediated lysis. Indeed, we demonstrated this to be the case using an indirect method for the assay of myocyte destruction, where continued expression of a luciferase reporter gene is used as a marker of transfected myocyte integrity.8 The vast majority of luciferase-expressing myocytes, which had been co-transfected with luciferase-expressing and antigen-expressing plasmids, were eliminated within 20 days of i.m. injection in normal immune competent mice, but not in SCID mice, indicating that the myocyte destruction was immune-mediated.8 Furthermore, in normal mice, the period of rapid myocyte destruction was accompanied by an influx of CD8+ and CD4+ T cells, as assessed by immunohistochemical staining.

In the present study, we have used mouse strains of various defined immunocompromised status in an attempt to determine more specifically which immunological mechanisms might be involved in the destruction of the antigen-expressing myocytes. Furthermore, in contrast to our previous study, the present study employs a co-linear expression system, where a single plasmid contains both luciferase- and HBsAg-encoding genes, each of which are under the control of their own cytomegaloviral (CMV) promoter (pCMV-S/CMV-luc). We demonstrate that the immune-mediated mycocyte destruction is due to mechanisms that are dependent on MHC II restricted CD4+ T cell activation but that are not mediated solely by MHC I-restricted or perforin-mediated lysis.

The implications of these results are important not only for DNA vaccines, but also for gene therapy applications where immune-mediated destruction of transfected cells may be undesirable. We have only evaluated myocytes in the present study since their structural post-mitotic nature allows meaningful long-term study, however their fate is likely representative of that for any transfected cell. As such, this system could be a valuable tool for evaluating the immunogenicity of gene-based vectors destined for gene therapy, regardless of the target tissue.

Results

The destruction of myocytes is associated with the expression of HBsAg

The pCMV-S/CMV-luc vector (Figure 1a) coding for both HBsAg and luciferase was injected i.m. in normal BABL/C (H-2d) and C57BL/6 mice (H-2b) to evaluate the rate of destruction of antigen-expressing muscle cells in fully immune-competent mice that served as congenic controls for the various immune-deficient strains of mice used in this study. Luciferase activity in the injected muscle was assessed at different time-points and expressed relative to the level found at 3 days. Activity was significantly decreased by 14 days in the BALB/c mice (Figure 2a) and by 21 days in the C57BL/6 mice (Figure 2c). Levels continued to decline such that by 42 days, luciferase activity was only 0.1% and 1% of 3 day levels in BALB/C and C57BL/6 mice, respectively. In contrast, with i.m. injection of the pCMV-(-S)/CMV-luc control vector (Figure 1b), which does not express HBsAg, a significant decrease in luciferase expression was only observed at 42 days, with the level of activity being 22% and 2% of 3-day levels in BALB/C and C57BL/6 mice respectively (Figure 2b and d).

Figure 1
figure1

DNA plasmids. The DNA plasmids used for the i.m. injections consisted of a HBsAg/luciferase co-linear expression plasmid (pCMV-S/CMV-luc) (a) which coded for both HBsAg and luciferase each under the control of their own CMV promoter and the control vector (pCMV-(S)/CMV-luc) (b) which contained the coding sequence for HBsAg but in the reverse orientation, therefore only luciferase was expressed from this construct.

Figure 2
figure2

The destruction of myocytes is associated with the expression of HBsAg. 50 μg of the HBsAg/luciferase co-linear expression vector (a and c) or control vector (b and d) were delivered i.m. to the TA muscle of normal BALB/c and C57BL/6 mice. The TA muscles (n = 10) were removed at different time-points after injection and the level of luciferase expression assessed by commercial assay. Error bars represent standard deviation. **P < 0.05; *P < 0.01.

The destruction of myocytes is immune-mediated

C.B-17 SCID (H-2d) mice fail to mount a functional adaptive immune response to specific antigen by virtue of a complete lack of B and T cells. Unlike the immune-competent mice, there was no significant decrease in luciferase activity over 42 days in the muscles of SCID mice that were injected with the HBsAg/luciferase colinear-expression vector (Figure 3a).

Figure 3
figure3

The destruction of myocytes is immune-mediated, but not solely due to an MHC 1 restricted, perforin-mediated lysis. 50 μg of the HBsAg/luciferase co-linear expression vector was delivered i.m. to the TA muscle of mice with different immunocompromised backgrounds: (a) SCID mice (H-2d), (b) perforin knockout mice (Pfn K/O) (H-2b), (c) MHC class I deficient mice (C1D) (H-2b), (d) MHC class II deficient mice (C2D) (H-2b). TA muscles (n = 10) were removed at different time-points after injection. Error bars represent standard deviation. *P < 0.01.

The destruction of myocytes is not solely due to an MHC I restricted, perforin-mediated lysis

Perforin knock-out (Pfn K/O, H-2b) mice are deficient in their ability to induce perforin-mediated lysis. In these mice, a decrease to about 2% (relative to the 3-day level) of luciferase expression was observed by 14 days after injection of the co-expression vector (Figure 3b). The kinetics and magnitude of this decrease were faster and stronger than those seen in normal congenic C57BL/6 H-2b control mice (Figure 2c).

C1D mice (also based on H-2b) lack functional β2-microglobulin, which is required for MHC class I molecules, resulting in an absence of mature CD8+ T cells.9 As such, these mice display a severe deficiency in any CD8+ T cell-mediated processes, whether mediated by perforin or not. The expression of luciferase from the co-expression vector was reduced over time in muscles of the C1D mice (Figure 3c), similar to that seen in the in the normal C57BL/6 H-2b control mice (Figure 2c), however the kinetics of the decrease was somewhat later and the magnitude somewhat less than that seen in the perforin-knockout mice (Figure 3b).

The destruction of myocytes requires MHC II expression

C2D mice (also based on H-2b) are knock-out mice that do not express any functional MHC class II molecules and do not have any mature CD4+ T cells,10 which are required for induction of both humoral and cell-mediated adaptive immune responses. In the absence of these immunological parameters, no significant decrease in luciferase activity in muscles injected with the co-expression vector was observed at any time-point, similar to the results in SCID mice (Figure 3d).

The destruction of myocytes is associated with HBsAg specific antibody production

In addition to evaluating the level of luciferase expression as a marker for muscle destruction, the production of HBsAg-specific antibodies was also determined. In all cases where there was a significant decrease in luciferase expression (associated with the expression of HBsAg), there was also a detectable HBsAg-specific antibody response (Figure 4a and b).

Figure 4
figure4

The destruction of myocytes is associated with HBsAg-specific antibody production. At 42 days after injection, plasma was collected from all groups of mice and their titre of HBsAg-specific IgG evaluated by ELISA in the presence or absence of HBsAg (a) and in presence HBsAg in different immunocomprimised backgrounds (b).

Discussion

We have confirmed our original observations that the loss of luciferase activity in muscles co-transfected with luciferase- and HBsAg-expressing vectors was associated with an antigen-specific immune-mediated destruction of transfected myocytes.8 In the present study, luciferase activity in muscles of immune-competent mice injected with a co-linear co-expression vector decreased significantly (>99%) by 14–21 days. In contrast, muscles of normal mice injected with a control vector expressing only luciferase did not have significantly decreased luciferase activity until 42 days, and even then the magnitude of the loss was much less that with the antigen-expressing vector.

In SCID mice (H-2d), no significant decrease in luciferase expression was observed over time, whether or not antigen was co-expressed. This suggests that the decrease in luciferase activity at 42 days in normal mice with the control vector was likely due to the induction of weak luciferase-specific immune responses, which took longer to develop than the HBsAg-specific responses. Although this was not observed in our previous study, a luciferase-specific immune response is supported in the literature.11 The differences in the kinetics and magnitude of muscle destruction between the BALB/C and the C57BL/6 mice is likely related to differences in H2-restricted recognition of T cell epitopes on the HBsAg and luciferase antigen.12

In muscles of perforin knock out mice (H-2b) injected with the co-expression vector, luciferase activity decreased sooner in the MHC-matched parental C57BL/6 strain of immune-competent mice. This finding was somewhat surprising since traditionally it has been accepted that CD8+ T cells mediate cellular destruction via a perforin- and granzyme-dependent induction of apoptosis.13 However, more recently, a CD95/CD95L-dependent mechanism of cytolytic activity has also been described.13 In addition to direct mechanisms of cellular lysis, CD8+ T cells can also secrete cytokines that have indirect effects including the inhibition of viral replication and the induction of apoptosis.13 Therefore, the results of the present study suggest that destruction of antigen-expressing myocytes following DNA vaccination occurs in part by perforin-independent mechanisms.

Luciferase activity also decreased over time in muscles of C1D mice that lack functional MHC I molecules and thus lack mature CD8+ T cells. The kinetics of this response were similar to those displayed by the MHC-matched parental C57BL/6 strain of immune-competent mice, however the magnitude was less. Destruction of antigen-expressing myocytes in this environment is likely due to antibody-dependent mechanisms of cellular lysis. In all cases where there was myocyte destruction, there was also the development of an HBsAg-specific antibody response. Since HBsAg is expressed as a transmembrane protein,14,15,16 two well-documented mechanisms of antibody-dependent cellular cytotoxicity (ADCC) could play an important role in HBsAg restricted cellular lysis, namely: (1) antibody-directed complement-mediated cytotoxicity;17 and (2) Fc receptor-mediated antibody-dependent cellular cytotoxicity.18

Therefore, in the present study, a perforin-mediated or CD8+ T cell/MHC I-restricted mechanism cannot solely explain antigen-specific immune-mediated destruction of myocytes.

As seen with the SCID mice, the immune-mediated destruction of HBsAg-expressing myocytes was completely abrogated in C2D mice that lack any functional MHC II molecules, and therefore also lack any mature CD4+ T cells. This underlines the crucial role of CD4+ T cells and MHC II-restricted professional antigen presentation in the development of a DNA vaccine-induced immune response.

There are two populations of APC that may be involved in the activation of the immune system following DNA-based vaccination. The first are APC that are directly transfected with the DNA plasmid and which both produce soluble antigen and present antigen in the context of MHC I and II. The second population are APC that are not directly transfected, yet are capable of antigen uptake either directly from the immediate environment or indirectly via Fc receptor-mediated uptake of antigen-antibody complexes, and in turn presenting it in the context of MHC I and II.19,20 Regardless of the scenario, both populations would constitute potential targets for MHC I restricted/CD8+ T cell-mediated lysis.21,22 The present results also suggest that directly transfected APCs could be susceptible to additional antibody-mediated mechanisms of cellular lysis based on the intracellular production and surface membrane expression of antigen.

The results of this study have important implications with respect to gene therapy. Unlike with DNA vaccines, immune responses against the expressed protein are generally not desired for gene therapy strategies. Overcoming unwanted immune responses has proven to be a very difficult task with successes only in the rare cases of genetic immune deficiencies.23 The observations made in the present study that transfected cells expressing foreign proteins are quite rapidly destroyed by several different immune mechanisms suggest that the restrictions imposed by the immune system on gene therapy may be more extensive than originally anticipated. However, the fact that luciferase expression by the control plasmid avoided stringent immune surveillance as compared with HBsAg expression, suggests that gene therapy may be possible for proteins that, by their very nature, are not efficiently recognized by the immune system. If these properties can be further defined, it may be possible to engineer gene therapy candidate proteins that are more immunologically neutral. These proteins could then be combined with transient immunosuppression or tolerization approaches to ensure prolonged transgene expression and more efficient gene therapy.24,25 As candidate proteins are discovered, the model systems presented herein could be used to screen their relative ability to stimulate an immune response.

Materials and methods

Mice

For immune-competent mice, 6–8-week-old female BALB/C and C57BL/6 mice (Charles River, Montreal, Canada) were used. Immune-compromised mice were bred in the transgenic facility of the Loeb Health Research Institute, Ottawa, Canada. Both male and female mice were used at 6–12 weeks of age and the breeding pairs were originally obtained from: severe combined immunodeficient (SCID) C.B-17 SCID mice (Charles River), major histocompatability complex (MHC) class I and class II deficient mice (Taconic, Germantown, NY, USA) and previously described perforin knock-out mice26 (Sandoz, Zurich, Switzerland).

Plasmid constructs

HBsAg/luciferase colinear-expression vector (pCMV-S/CMV-luc):

A 3.4 kb SmaI–BamHI fragment containing the luciferase reporter gene, under control of the immediate–early cytomegalovirus (CMV) promoter was isolated from the previoulsy described pCMV-luc vector.27 The SmaI–BamHI fragment was gel purified, the 5′ overhang filled in by DNA polymerase I large fragment, and cloned into the StuI site of the previously described pMAS-S vector,28 which contains the coding sequence for the HBsAg. The resulting HBsAg/luciferase colinear-expression vector (pCMV-S/CMV-luc) (Figure 1a) coded for both HBsAg and luciferase each under the control of their own CMV promoter.

The control vector (pCMV-(-S)/CMV-luc):

A 692 bp EcoRV–PstI fragment containing the HBsAg gene was isolated from the previously described pMAS-S vector28 and cloned into a PstI–SmaI digested pMAS vector that did not contain the HBsAg gene. This cloning strategy resulted in the insertion of the HBsAg gene into the pMAS vector in the reverse orientation yielding the pMAS-(-S) control vector. Subsequently, the 3.4 kb SmaI–BamHI CMV-luc fragment was cloned into pMAS-(-S), as described above, and the resulting pCMV-(-S)/CMV-luc vector (Figure 1b) expressed luciferase, but not HBsAg.

Plasmid purification

The plasmid DNA was purified on Qiagen anion-exchange chromatography columns (Qiagen, Hilden, Germany) and redissolved in 0.15 M NaCl. The DNA was stored at −20°C until required for direct injection into the TA muscles of mice.

Indirect assay of muscle fiber destruction:

Mice were injected bilaterally with 50 μg of the pCMV-S/CMV-luc vector (all strains) or the control pCMV-(-S)/CMV-luc (BALB/c and C57BL/6 mice only) in the TA muscles, as described previously.29 Groups of animals (n = 5) of a given strain that were injected with a given plasmid DNA, were anesthetized and killed by cervical dislocation at 3, 7, 14, 21 and 42 days after injection. The TA muscles were removed and luciferase activity was measured using the Promega luciferase assay system (Promega, Madison, WI, USA) as previously described.27 Results were expressed as group means ± s.d. of relative light units (RLU)/s/mg protein, with protein contents being determined by the BioRad microassay procedure (BioRad, Mississauga, ON, Canada).30

Assay of humoral response to HBsAg:

Titers of antibodies (Ab) specific to HBsAg (anti-HBs) were determined in plasma recovered from mice at 42 days after injection by enzyme-linked immunosorbent assay (ELISA) as previously described.31 Anti-HBs titers were defined as the highest plasma dilution that resulted in an absorbance (OD 450) two times greater than that of non-immune plasma. Titers were expressed as the geometric mean ± s.d. of values from individual animals in a treatment group, which were themselves the average of at least duplicate assays.

Statistical analysis

The statistical significance of differences between group mean luciferase activity in muscles taken from a given mouse strain injected with a given vector at various times after DNA injection was determined by one-factor analysis of variance followed by the Dunnett's multiple comparison test. Differences of values at other time-points relative to the 3-day activity were considered significant with a P 0.05.

References

  1. 1

    Davis HL, Michel ML, Whalen RG . DNA-based immunization induces continuous secretion of hepatitis B surface antigen and high levels of circulating antibody Hum Molec Genet 1993 2: 1847–1851

  2. 2

    Davis HL et al. Direct gene transfer in skeletal muscle: plasmid DNA-based immunization against the hepatitis B virus surface antigen Vaccine 1994 12: 1503–1509

  3. 3

    Davis HL, Schirmbeck R, Reimann J, Whalen RG . DNA-mediated immunization in mice induces a potent MHC class I-restricted cytotoxic T lymphocyte response to the hepatitis B envelope protein Hum Gene Ther 1995 6: 1447–1456

  4. 4

    Kowalczyk DW, Ertl HC . Immune responses to DNA vaccines Cell Mol Life Sci 1999 55: 751–770

  5. 5

    Davis HL, Mancini M, Michel ML, Whalen RG . DNA-mediated immunization to hepatitis B surface antigen: longevity of primary response and effect of boost Vaccine 1996 14: 910–915

  6. 6

    Wolff JA et al. Long-term persistence of plasmid DNA and foreign gene expression in mouse muscle Hum Mol Genet 1992 1: 363–369

  7. 7

    Cheng L, Ziegelhoffer PR, Yang NS . In vivo promoter activity and transgene expression in mammalian somatic tissues evaluated by using particle bombardment Proc Natl Acad Sci USA 1993 90: 4455–4459

  8. 8

    Davis HL, Millan CL, Watkins SC . Immune-mediated destruction of transfected muscle fibers after direct gene transfer with antigen-expressing plasmid DNA Gene Therapy 1997 4: 181–188

  9. 9

    Zijlstra M et al. Beta 2-microglobulin deficient mice lack CD4–8+ cytolytic T cells Nature 1990 344: 742–746

  10. 10

    Grusby MJ, Johnson RS, Papaioannou VE, Glimcher LH . Depletion of CD4+ T cells in major histocompatibility complex class II-deficient mice Science 1991 253: 1417–1420

  11. 11

    Klavinskis LS, Barnfield C, Gao L, Parker S . Intranasal immunization with plasmid DNA–lipid complexes elicits mucosal immunity in the female genital and rectal tracts J Immunol 1999 162: 254–262

  12. 12

    Schirmbeck R et al. Nucleic acid vaccination primes hepatitis B virus surface antigen-specific cytotoxic T lymphocytes in nonresponder mice J Virol 1995 69: 5929–5934

  13. 13

    Harty JT, Tvinnereim AR, White DW . CD8+ T cell effector mechanisms in resistance to infection Annu Rev Immunol 2000 18: 275–308

  14. 14

    Gholson CF, Siddiqui A, Vierling JM . Cell surface expression of hepatitis B surface and core antigens in transfected rat fibroblast cell lines Gastroenterology 1990 98: 968–975

  15. 15

    Chu CM, Liaw YF . Membrane staining for hepatitis B surface antigen on hepatocytes: a sensitive and specific marker of active viral replication in hepatitis B J Clin Pathol 1995 48: 470–473

  16. 16

    Satoh O et al. Membrane structure of the hepatitis B virus surface antigen particle J Biochem (Tokyo) 2000 127: 543–550

  17. 17

    Michalak TI et al. Antibody-directed complement-mediated cytotoxicity to hepatocytes from patients with chronic hepatitis B Clin Exp Immunol 1995 100: 227–232

  18. 18

    Daeron M . Fc receptor biology Annu Rev Immunol 1997 15: 203–234

  19. 19

    Machy P, Serre K, Leserman L . Class I-restricted presentation of exogenous antigen acquired by Fcgamma receptor-mediated endocytosis is regulated in dendritic cells Eur J Immunol 2000 30: 848–857

  20. 20

    Mitchell DA, Nair SK, Gilboa E . Dendritic cell/macrophage precursors capture exogenous antigen for MHC class I presentation by dendritic cells (published erratum appears in Eur J Immunol 1998; 28: 3891) Eur J Immunol 1998 28: 1923–1933

  21. 21

    Knight SC, Askonas BA, Macatonia SE . Dendritic cells as targets for cytotoxic T lymphocytes Adv Exp Med Biol 1997 417: 389–394

  22. 22

    Parajuli P et al. Cytolysis of human dendritic cells by autologous lymphokine-activated killer cells: participation of both T cells and NK cells in the killing J Leukoc Biol 1999 65: 764–770

  23. 23

    Cavazzana-Calvo M et al. Gene therapy of human severe combined immunodeficiency (SCID)-X1 disease Science 2000 28: 669–672

  24. 24

    Wilson CB, Embree LJ, Schowalter D et al. Transient inhibition of CD28 and CD40 ligand interactions prolongs adenovirus-mediated transgene expression in the lung and facilitates expression after secondary vector administration J Virol 1998 72: 7542–7550

  25. 25

    Halbert CL, Standaert TA, Wilson CB, Miller AD . Successful readministration of adeno-associated virus vectors to the mouse lung requires transient immunosuppression during the initial exposure J Virol 1998 72: 9795–9805

  26. 26

    Kagi D et al. Cytotoxicity mediated by T cells and natural killer cells is greatly impaired in perforin-deficient mice Nature 1994 369: 31–37

  27. 27

    Davis HL, Whalen RG, Demeneix BA . Direct gene transfer into skeletal muscle in vivo: factors affecting efficiency of transfer and stability of expression Hum Gene Ther 1993 4: 151–159

  28. 28

    Krieg AM et al. Sequence motifs in adenoviral DNA block immune activation by stimulatory CpG motifs Proc Natl Acad Sci USA 1998 95: 12631–12636

  29. 29

    Davis HL et al. Plasmid DNA is superior to viral vectors for direct gene transfer into adult mouse skeletal muscle Hum Gene Ther 1993 4: 733–740

  30. 30

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

  31. 31

    Davis HL et al. CpG DNA is a potent enhancer of specific immunity in mice immunized with recombinant hepatitis B surface antigen J Immunol 1998 160: 870–876

Download references

Acknowledgements

We wish to thank Jocelyn Chandler, Brian Avery, George Condruit, and Chandimal Nicholas for their excellent technical support. This work was supported by a grant from the Medical Research Council of Canada to HLP. PJ Payette is a recipient of an Ontario Graduate Student Award and HL Davis is the recipient of an Ontario Ministry of Health Career Scientist Award.

Author information

Correspondence to HL Davis.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Payette, P., Weeratna, R., McCluskie, M. et al. Immune-mediated destruction of transfected myocytes following DNA vaccination occurs via multiple mechanisms. Gene Ther 8, 1395–1400 (2001). https://doi.org/10.1038/sj.gt.3301534

Download citation

Keywords

  • DNA vaccines
  • myocytes
  • reporter genes
  • cytolytic destruction

Further reading