- Email: [email protected]
RNA degradation precedes DNA cleavage in autoreactive CD4 T cells suppressed by calicheamicin γ1
International Immunopharmacology 4 (2004) 521 – 526 www.elsevier.com/locate/intimp
RNA degradation precedes DNA cleavage in autoreactive CD4 T cells suppressed by calicheamicin g1 Sunil Thomas, Anca Preda-Pais, Sofia Casares, Teodor-D. Brumeanu * Mount Sinai School of Medicine, Department of Microbiology, 1 Gustave L. Levy Place, New York, NY 10029, USA Received 20 November 2003; received in revised form 12 December 2003; accepted 6 January 2004
Abstract Calicheamicin g1 (Cal g1) is a hydrophobic enediyne antibiotic known to cleave the DNA and lead to apoptosis in a variety of cells. Herein, we show that Cal g1 exhibits a 1000-times stronger suppressogenic effect on antigen-specific (diabetogenic), and naı¨ve CD4 T cells than Doxorubicin (Dox), another strong apoptotic drug. The thymic precursors and mature T cells incubated with Cal g1 for only 30 min showed a drastic decrease or loss of cytokine production and proliferation following stimulation with the immunogenic peptide, or with CD3 and CD28 antibodies. The suppressogenicity of Cal g1 correlated with a rapid and non-selective degradation of RNA, whereas the DNA cleavage occurred at a later time point and at higher doses. Cal g1 may represent a potential therapeutic agent to eliminate self-reactive T cells in autoimmune diseases, providing that is delivered by antigen-specific T-cell ligands. Targeting of highly suppressogenic drugs such as Cal g1 to autoreactive T cells may reduce considerable the therapeutic dose and the drug-related side effects. D 2004 Elsevier B.V. All rights reserved. Keywords: Calicheamicin g1; Autoreactive CD4 T cells; RNA degradation; Cell death
1. Introduction Cal g1 is a hydrophobic enediyne antibiotic secreted by Micromonospora echinospora ssp. calichensis [1]. The drug was shown to cause severe DNA damage by binding to the minor groove of the DNA helix, and removal of hydrogen atoms from deoxyribose with subsequent cleavage of the double stranded DNA [2]. Several Cal g1-antibody conjugates specific for B-cell antigens such as CD33 and * Corresponding author. Tel.: +1-212-241-7551; fax: +1-212423-0711. E-mail address: [email protected] (T.-D. Brumeanu). 1567-5769/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.intimp.2004.01.002
CD22 showed promising therapeutic effects in Bcell malignancies [3,4]. At present, Gemtuzumab Ozogamicin (Mylotarg), a commercial preparation made of Cal g1 chemically coupled to a CD33 antibody, is the only Cal g1-based chemotherapeutic successfully used for the therapy of acute myeloid leukemia. Drug-mediated apoptosis of unwanted cells such as tumor cells and self-reactive immune cells is an attractive strategy aimed at eliminating these cells from the system. Several reports showed the possibility of targeting tumor cells with antimitotic drugs and toxins delivered through specific antibodies. We showed that a CEA-Dox conjugate antibody can kill colon cancer cells in vivo [5]. Drug targeting of self-
522
S. Thomas et al. / International Immunopharmacology 4 (2004) 521–526
reactive T cells in autoimmune diseases is a relatively new concept. We have previously reported that in vivo targeting of Dox to peptide-specific CD4 T cells through a physiological ligand for T cells, namely a soluble dimeric peptide-MHC II chimera (DEF), can efficiently kill thymic precursors and mature diabetogenic T cells in vivo [6]. Herein, we compared the anti-mitogenic (suppressogenic) capacity of Cal g1 and Dox-free drugs on the antigen-specific (diabetogenic) T cells and naı¨ve CD4 T cells in vitro. We found that Cal g1 exhibited a 1000-times stronger suppressogenic effect on T cells than Dox. The mechanism underlying Cal g1 suppressogenicity appears to rely primarily on a rapid and non-selective RNA degradation before the DNA cleavage occurs.
2. Materials and methods 2.1. Mice Balb/c mice were purchased from Jackson Labs, ME, and they were used as a source for naı¨ve CD4 T cells. The Balb/c mice expressing the 14.3-day (transgenic, Tg) T-cell receptor (TCR) specific for the hemagglutinin peptide HA110-120 of PR8 influenza virus (TCR-HA Tg mice) were previously described [7], and they were used in these experiments as a source for antigen-specific (diabetogenic) T cells. All mice were maintained at Mount Sinai School of Medicine in a pathogen-free facility according to federal and local regulations. 2.2. Cells The thymic precursors and mature, HA-specific CD4 T cells and naı¨ve CD4 T cells were negatively sorted from the thymus and spleen of 1-month-old TCR-HA Tg mice and non-immunized Balb/c mice, respectively, using mouse CD4 column kits (R&D, MN). We previously showed that the HA-specific CD4 T cells from TCR-HA Tg mice are diabetogenic when infused in RAG2 deficient RIP-HA Tg mice, or hosted in TCR-HA+/ , RIP-HA+/ double transgenic mice [7]. The antigen presenting cells (APC) were prepared as plastic adherent cells from the spleen of TCR-HA Tg mice or naı¨ve Balb/c mice.
2.3. Drugs Cal g1 (Wyeth, NY) and Dox (Aldrich, NJ) were solubilized in ethanol and DMSO respectively, as a stock solution (10 mg/ml), and kept frozen till used. 2.4. Thymidine incorporation assay ( 3H-TdR) Thymocytes or mature splenic T cells (5 105 cells) from TCR-HA Tg mice in the presence of HA110-120 synthetic peptide (10 Ag/ml) and 2200 rads-irradiated APCs (2 105 cells) were incubated in RPMI 1640 medium (containing 10% fetal calf serum, FCS) for 3 days in 96-well Costar plates at 37 jC with graded amounts of Cal g1 or Dox-free drugs. The tritiated tymidine (1 ACi/well) was added to the cultures for the last 24 h, cells were harvested on Skatron filter paper, and the radioactivity (cpm) was measured in triplicate wells using a h-scintillation chamber. In parallel assays, negatively sorted, naı¨ve CD4 T cells from Balb/c mice were stimulated in a non-antigen-specific manner using a mixture of CD3 (2C11 mAb) and CD28 monoclonal antibodies (BD PharMingen, CA). Stimulation with CD3/CD28 Abs induces the same index of proliferation and pattern of cytokine secretion in both the TCR-HA and naı¨ve CD4 T cells, and it was much similar with the stimulation with HA-pulsed APCs. To avoid the interference with RNA and DNA from APCs in gel electrophoresis and RT-PCR, the stimulation of T cells with CD3/CD28 Abs was the system of choice. 2.5. Cytokine assay The cytokine production was measured by ELISA (mouse ELISA kits, Biosource International, CA) in the cell culture supernatants from parallel cultures of thymidine incorporation assay upon stimulation with HA peptide-pulsed APCs or a mixture of CD3 and CD28 antibodies. 2.6. DNA and RNA analysis DNAwas extracted from splenic TCR-HATg T cells using the conventional ethanol-based purification method, and total RNA using NucleoSpin RNA II kit (BD Biosciences Clontech, CA). The DNA and RNA preparations (4 Ag) were incubated for various intervals
S. Thomas et al. / International Immunopharmacology 4 (2004) 521–526
523
Fig. 1. Inhibition of proliferation of CD4 antigen-specific and non-antigen-specific T cells by Cal g1 and Dox drugs. Negatively sorted CD4 from TCR-HA Tg mice and naı¨ve Balb/c mice were cultured for 3 days with HA-pulsed, irradiated APCs (left panel) or CD3 and CD28 antibodies (right panel), respectively, in the presence of graded amounts of drugs. The 3H-thymidine was added for the last 24 h, and thymidine incorporation was calculated as described. The horizontal dotted lines represent the 50% index of T-cell proliferation, and the vertical dotted lines indicate the corresponding drug concentration. The index of proliferation of CD4 TCR-HA T cells stimulated with HA-pulsed APCs was 82,000 F 950 cpm, and that of naı¨ve CD4 T cells stimulated with CD3/CD28 Abs in the absence of Cal g1 was 60,500 F 1150 cpm.
of time with graded amounts of drugs, and analyzed by electrophoresis in 1.5% agarose gel containing ethidium bromide. For the semi-quantitative RT-PCR analysis of IL-2, IL-4, IL-10 and IFN-g gene transcription, the cDNA was prepared from cells treated with Cal g1 using the Qiagen One Step RT-PCR kit (Qiagen Valencia, CA) and the following specific primers: IL-2
(forward GACACTTGTGCTCCTTGTCA, reverse TCAATTCTGTGGCCTGCTTG), IL-4 (forward TCGGCATTTTGAACGAGGTC, reverse GAAAAGCCCGAAAGAGT CTC), IL-10 (forward 5VATGCAGGACTTTAAGGGTTACTTG, reverse TAGACACCTTG GTCTTGGAGCTTA) and IFN-g (forward GCTCTGAGACAATGAACGCT, reverse
Table 1 Inhibition of cytokine production in T cells by Cal g1 Cytokine production (pg/ml F S.D.)
HA/APC + TCR-HA CD4 T cells + Nil CD3/CD28 Ab + naı¨ve CD4 T cells + Nil Cal g1 (3 ng/ml for 30 min)a + HA/APC + TCR-HA CD4 T cells Cal g1 (3 ng/ml for 30 min)a + CD3/CD28 Ab + naı¨ve CD4 T cells Cal g1 (3 ng/ml for 2 h)a + HA/APC + TCR-HA CD4 T cells Cal g1 (3 ng/ml for 2 h)a + CD3/CD28 Ab + naı¨ve CD4 T cells TCR-HA CD4 T cells + medium alone (background) Naı¨ve CD4 T cells + medium alone (background) a
IL-2
IFN-g
IL-4
IL-10
780 F 3.5 615 F 4.0 280 F 2.0
960 F 2.5 998 F 3.1 48 F 1.8
405 F 1.0 115 F 2.3 10.8 F 1.4
97 F 1.5 12.5 F 2.1 21 F 3.7
141 F 1.7
64 F 2.0
12.2 F 1.0
< 5.5
32 F 3.4
21 F 2.8
10.0 F 2.0
< 5.5
< 5.5
< 5.5
< 5.5
48 F 4.1
37 F 3.1
12 F 2.4
7.5 F 1.8
27 F 3.7
30 F 1.8
7.5 F 2.0
< 5.5
17.5 F 3.1
T cells were pre-treated with Cal g1 for 30 min or 2 h, washed and cultured for 3 days in drug-free medium with the stimulatory HA110120 peptide and irradiated APCs (in the case of TCR-HA CD4 T cells from transgenic mice), or with CD3 plus CD28 stimulatory antibodies (in the case of naı¨ve CD4 T cells from Balb/c mice). Cytokines were measured by ELISA. Shown are the values (pg/ml) from triplicate wells F S.D.
524
S. Thomas et al. / International Immunopharmacology 4 (2004) 521–526
of HA110-120-specific (diabetogenic) CD4 T cells and naı¨ve CD4 T cells. Only 50% of cells proliferated upon continuous incubation with 3 ng/ml of Cal g1, whereas the same effect was obtained with 3000 ng/ ml of Dox. At 100 ng/ml of Cal g1, the HA proliferative response of splenic TCR-HA T cells from Tg mice, and the CD3/CD28-Ab-mediated proliferation of naı¨ve CD4 T cells from Balb/c mice, was abrogated (Fig. 1). The TCR-HA and naı¨ve CD4 thymic T-cell precursors showed similar sensitivity to Cal g1 as the mature T cells (data not shown). The fact that only 50% of T cells ceased proliferation upon a 3-day continuous incubation with 3 ng/ml of Cal g1 suggested that not all cells can uptake the optimal suppressogenic amount of the drug. This raised the question of whether a certain extent of drug degradation may occur in the culture medium. The uptake of Cal g1 by splenic T cells occurred relatively fast, since 30 min pre-incubation with 12.5 ng/ml Cal g1 ceased the HA proliferative response by 80%, and higher doses ceased completeFig. 2. Cal g1 induces fast and severe RNA degradation in TCR-HA T cells. Panel A: the RNA and DNA was extracted from negatively sorted, splenic CD4 TCR-HA T cells, and incubated on ice for various intervals of time and doses, as indicated. The 28s and 18s major RNA bands, and the DNA band were analyzed for the integrity by 1.5% agarose gel electrophoresis. Some 4 Ag/lane of total RNA and DNA material was analyzed. Shown is one of three representative experiments. Panel B: the CD4 TCR-HA T cells (108 cells) were incubated for various intervals of time and doses of Cal g1 as indicated, and the cells were washed and cultured in drug-free medium for another 16 h. The total RNA and DNA was extracted and analyzed as described. Shown is one of two representative experiments carried out with 20 Ag/lane of nucleic material.
AAAGA GATAATCTGGCTGTGC) [8]. The PCR products were separated by electrophoresis in 1.5% agarose gel containing ethidium bromide.
3. Results and discussion 3.1. Suppressogenicity of Cal c1 on T cells is f1000fold higher than that of Dox The index of cell proliferation determined by 3HTdR assay indicated that Cal g1 was f 1000-fold more potent than Dox in suppressing the cell growth
Fig. 3. Cal g1 induces non-selective degradation of interleukins mRNA in TCR-HA T cells. Negatively sorted, splenic CD4 TCRHA T cells were incubated or not for 30 min and 2 h with 5 ng/ml Cal g1, cells were washed and cultured for 16 h with CD3/CD28 stimulatory antibodies. Cells were washed, the total RNA was extracted and cDNA prepared as described. Some 4 Ag/lane of cDNA material was analyzed by RT-PCR using specific primers for IL-2, IL-4, IL-1L and IFN-g. Shown is the amplicon size (base pairs, bp) for each interleukin from one of three representative experiments. Lane 1: cells cultured with medium alone for 16 h in the absence of Cal g1 (control, signal-to-noise background); lane 2: cells stimulated with CD3/CD28 Abs for 16 h in the absence of Cal g1 (control stimulation); lane 3: cells treated for 30 min with Cal g1 (5 ng/ml) and stimulated for 16 h with CD3/CD28 Abs in drug free medium; and lane 4: cells treated for 2 h with Cal g1 (5 ng/ml) and stimulated for 16 h with CD3/CD28 Abs in drug-free medium.
S. Thomas et al. / International Immunopharmacology 4 (2004) 521–526
ly the proliferation. A bystander toxic effect of Cal g1 on APCs has been ruled out in these experiments, since the T cells were first incubated with the drug, and then washed, and incubated with healthy APCs pulsed with HA peptide. These results indicated that Cal g1 toxicity relies mostly on its cellular uptake, and less on its chemical stability. At present, neutralization of Cal g1 toxicity by binding to serum proteins can not be ruled out. This remains to be further investigated. 3.2. Suppressogenicity of Cal c1 on T cells relies mainly on RNA degradation The cell culture supernatants from a 3-day 3H-TdR assay carried out in the presence of toxic doses of Cal g1 showed a lack of IL-2, IFN-g, IL-4 and IL-10 secretion. Similar results were obtained in the case of naı¨ve CD4 T cells from Balb/c mice during stimulation with CD3 and CD28 antibodies (Table 1). This correlated with a profound suppression of the HA proliferative response or CD3/CD28-induced proliferative response, as described (Fig. 1). Analysis of RNA and DNA integrity in TCR-HA T cells suppressed by Cal g1 showed a severe degradation of RNA 30 min after incubation with 5 ng/ml Cal g1, whereas the DNA cleavage occurred 2 h later and at a 10-times higher dose (50 ng/ml) (Fig. 2). The RNA and DNA degradation patterns in naı¨ve CD4 T cells from Balb/c mice were similar to those from TCR-HA T cells (data not shown). It is well known that IL-2 and IL-4 supports the growth of Th1 and Th2 cell subsets, respectively [9]. As the total RNA was rapidly degraded in T cells suppressed by Cal g1, we next analyzed their levels of IL-2, IL-4, IL-10 and IFN-g mRNA expression by semi-quantitative RT-PCR (Fig. 3). Not surprisingly, the mRNA expression for these cytokines was drastically inhibited after 30 min, and completely inhibited after 2 h of incubation with 5 ng/ml of Cal g1. This clearly indicated that a rapid and non-selective RNA degradation was the major cause of Cal g1-induced suppression of cytokine production with subsequent abrogation of cell growth. The mechanism of Cal g1mediated degradation of RNA maybe similar to that described for DNA [2], but far more efficient. This is because RNA is built as a single strand macromolecule, and DNA as a double strand macromolecule.
525
Obviously, a single-strand macromolecule could be easily accessed by the drug. Several reports indicated that RNA degradation is an earlier and more efficient mechanism of cell growth arrest and apoptosis than the DNA degradation [10]. In conclusion, our results demonstrated for the first time that Cal g1 is an extremely potent killing agent for both antigen-specific and non-specific CD4 T cells by a mechanism relying primarily on RNA degradation. The DNA cleavage occurred at higher doses of Cal g1 and after longer time of interaction with the drug. We have previously demonstrated that apoptotic drugs like Doxorubicin can efficiently eliminate diabetogenic T cells in vivo when delivered through a soluble, dimeric peptide-MHC II chimeras (DEF) [6]. Accordingly, one may consider that immunoconjugates of Cal g1 and DEF-like chimeras could provide rational grounds to develop new classes of therapeutic drugs for autoimmune diseases. DEF-Cal g1 conjugates may reduce considerable the therapeutic dose and number of injections, and the systemic toxic effects of the drug in vivo.
Acknowledgements This work was supported by grants from the National Institutes of Health (DK61927 and DK61326) to T.-D.B.
References [1] Love SF, Maiese WM, Rothstein DM. Conditions for protoplasting, regenerating, and transforming the calicheamicin producer, Micromonospora echinospora. Appl Environ Microbiol 1992;58:1376 – 8. [2] Nicolaou KC, Pitsinos N, Theodorakis EA, Saimoto H, Wrasidlo W. Synthetic calicheamicin mimics with novel initiation mechanisms: DNA cleavage, cytotoxicity, and apoptosis. Chem Biol 1994;1:57 – 66. [3] DiJoseph JF, Armellino DC, Boghaert ER, Khandke K, Dougher MM, Sridharan L, et al. Antibody-targeted chemotherapy with CMC-544: a CD22-targeted immunoconjugate of calicheamicin for the treatment of B lymphoid malignancies. Blood 2004;103 (in press). [4] Hamann PR, Hinman LM, Hollander I, Beyer CF, Lindh D, Holcomb R, et al. Gemtuzumab ozogamicin, a potent and selective anti-CD33 antibody-calicheamicin conjugate for treatment of acute myeloid leukemia. Bioconjug Chem 2002; 13:47 – 58.
526
S. Thomas et al. / International Immunopharmacology 4 (2004) 521–526
[5] Stan CA, Radu DL, Casares S, Bona C, Brumeanu TD. Antineoplastic efficacy of Doxorubicin enzymatically assembled on galactose residues of a monoclonal antibody specific for the carcinoembryonic antigen (CEA). Cancer Res 1999;59:115 – 21. [6] Casares S, Bona C, Brumeanu TD. Antigen-specific downregulation of T-cells by doxorubicin delivered by MHC IIpeptide chimeras. Nat Biotechnol 2001;19:142 – 7. [7] Casares S, Hurtado A, von Boehmer H, Sarukhan A, McEvoy R, Brumeanu TD. Downregulation of diabetogenic T-cells by a soluble MHC II-peptide chimera. Nat Immunol 2002; 3:383 – 91.
[8] Zhou HR, Yan D, Pestka JJ. Differential cytokine mRNA expression in mice after oral exposure to the trichothecene vomitoxin (deoxynivalenol):dose response and time course. Toxicol Appl Pharmacol 1997;144:294 – 305. [9] Watowich SS, Wu H, Socolovsky M, Klingmuller U, Constantinescu SN, Lodish HF. Cytokine receptor signal transduction and the control of hematopoietic cell development. Annu Rev Cell Dev Biol 1996;12:91 – 128. [10] Tourriere H, Chebli K, Tazi J. mRNA degradation machines in eukaryotic cells. Biochimie 2002;84:821 – 37.
RNA degradation precedes DNA cleavage in autoreactive CD4 T cells suppressed by calicheamicin g1 Sunil Thomas, Anca Preda-Pais, Sofia Casares, Teodor-D. Brumeanu * Mount Sinai School of Medicine, Department of Microbiology, 1 Gustave L. Levy Place, New York, NY 10029, USA Received 20 November 2003; received in revised form 12 December 2003; accepted 6 January 2004
Abstract Calicheamicin g1 (Cal g1) is a hydrophobic enediyne antibiotic known to cleave the DNA and lead to apoptosis in a variety of cells. Herein, we show that Cal g1 exhibits a 1000-times stronger suppressogenic effect on antigen-specific (diabetogenic), and naı¨ve CD4 T cells than Doxorubicin (Dox), another strong apoptotic drug. The thymic precursors and mature T cells incubated with Cal g1 for only 30 min showed a drastic decrease or loss of cytokine production and proliferation following stimulation with the immunogenic peptide, or with CD3 and CD28 antibodies. The suppressogenicity of Cal g1 correlated with a rapid and non-selective degradation of RNA, whereas the DNA cleavage occurred at a later time point and at higher doses. Cal g1 may represent a potential therapeutic agent to eliminate self-reactive T cells in autoimmune diseases, providing that is delivered by antigen-specific T-cell ligands. Targeting of highly suppressogenic drugs such as Cal g1 to autoreactive T cells may reduce considerable the therapeutic dose and the drug-related side effects. D 2004 Elsevier B.V. All rights reserved. Keywords: Calicheamicin g1; Autoreactive CD4 T cells; RNA degradation; Cell death
1. Introduction Cal g1 is a hydrophobic enediyne antibiotic secreted by Micromonospora echinospora ssp. calichensis [1]. The drug was shown to cause severe DNA damage by binding to the minor groove of the DNA helix, and removal of hydrogen atoms from deoxyribose with subsequent cleavage of the double stranded DNA [2]. Several Cal g1-antibody conjugates specific for B-cell antigens such as CD33 and * Corresponding author. Tel.: +1-212-241-7551; fax: +1-212423-0711. E-mail address: [email protected] (T.-D. Brumeanu). 1567-5769/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.intimp.2004.01.002
CD22 showed promising therapeutic effects in Bcell malignancies [3,4]. At present, Gemtuzumab Ozogamicin (Mylotarg), a commercial preparation made of Cal g1 chemically coupled to a CD33 antibody, is the only Cal g1-based chemotherapeutic successfully used for the therapy of acute myeloid leukemia. Drug-mediated apoptosis of unwanted cells such as tumor cells and self-reactive immune cells is an attractive strategy aimed at eliminating these cells from the system. Several reports showed the possibility of targeting tumor cells with antimitotic drugs and toxins delivered through specific antibodies. We showed that a CEA-Dox conjugate antibody can kill colon cancer cells in vivo [5]. Drug targeting of self-
522
S. Thomas et al. / International Immunopharmacology 4 (2004) 521–526
reactive T cells in autoimmune diseases is a relatively new concept. We have previously reported that in vivo targeting of Dox to peptide-specific CD4 T cells through a physiological ligand for T cells, namely a soluble dimeric peptide-MHC II chimera (DEF), can efficiently kill thymic precursors and mature diabetogenic T cells in vivo [6]. Herein, we compared the anti-mitogenic (suppressogenic) capacity of Cal g1 and Dox-free drugs on the antigen-specific (diabetogenic) T cells and naı¨ve CD4 T cells in vitro. We found that Cal g1 exhibited a 1000-times stronger suppressogenic effect on T cells than Dox. The mechanism underlying Cal g1 suppressogenicity appears to rely primarily on a rapid and non-selective RNA degradation before the DNA cleavage occurs.
2. Materials and methods 2.1. Mice Balb/c mice were purchased from Jackson Labs, ME, and they were used as a source for naı¨ve CD4 T cells. The Balb/c mice expressing the 14.3-day (transgenic, Tg) T-cell receptor (TCR) specific for the hemagglutinin peptide HA110-120 of PR8 influenza virus (TCR-HA Tg mice) were previously described [7], and they were used in these experiments as a source for antigen-specific (diabetogenic) T cells. All mice were maintained at Mount Sinai School of Medicine in a pathogen-free facility according to federal and local regulations. 2.2. Cells The thymic precursors and mature, HA-specific CD4 T cells and naı¨ve CD4 T cells were negatively sorted from the thymus and spleen of 1-month-old TCR-HA Tg mice and non-immunized Balb/c mice, respectively, using mouse CD4 column kits (R&D, MN). We previously showed that the HA-specific CD4 T cells from TCR-HA Tg mice are diabetogenic when infused in RAG2 deficient RIP-HA Tg mice, or hosted in TCR-HA+/ , RIP-HA+/ double transgenic mice [7]. The antigen presenting cells (APC) were prepared as plastic adherent cells from the spleen of TCR-HA Tg mice or naı¨ve Balb/c mice.
2.3. Drugs Cal g1 (Wyeth, NY) and Dox (Aldrich, NJ) were solubilized in ethanol and DMSO respectively, as a stock solution (10 mg/ml), and kept frozen till used. 2.4. Thymidine incorporation assay ( 3H-TdR) Thymocytes or mature splenic T cells (5 105 cells) from TCR-HA Tg mice in the presence of HA110-120 synthetic peptide (10 Ag/ml) and 2200 rads-irradiated APCs (2 105 cells) were incubated in RPMI 1640 medium (containing 10% fetal calf serum, FCS) for 3 days in 96-well Costar plates at 37 jC with graded amounts of Cal g1 or Dox-free drugs. The tritiated tymidine (1 ACi/well) was added to the cultures for the last 24 h, cells were harvested on Skatron filter paper, and the radioactivity (cpm) was measured in triplicate wells using a h-scintillation chamber. In parallel assays, negatively sorted, naı¨ve CD4 T cells from Balb/c mice were stimulated in a non-antigen-specific manner using a mixture of CD3 (2C11 mAb) and CD28 monoclonal antibodies (BD PharMingen, CA). Stimulation with CD3/CD28 Abs induces the same index of proliferation and pattern of cytokine secretion in both the TCR-HA and naı¨ve CD4 T cells, and it was much similar with the stimulation with HA-pulsed APCs. To avoid the interference with RNA and DNA from APCs in gel electrophoresis and RT-PCR, the stimulation of T cells with CD3/CD28 Abs was the system of choice. 2.5. Cytokine assay The cytokine production was measured by ELISA (mouse ELISA kits, Biosource International, CA) in the cell culture supernatants from parallel cultures of thymidine incorporation assay upon stimulation with HA peptide-pulsed APCs or a mixture of CD3 and CD28 antibodies. 2.6. DNA and RNA analysis DNAwas extracted from splenic TCR-HATg T cells using the conventional ethanol-based purification method, and total RNA using NucleoSpin RNA II kit (BD Biosciences Clontech, CA). The DNA and RNA preparations (4 Ag) were incubated for various intervals
S. Thomas et al. / International Immunopharmacology 4 (2004) 521–526
523
Fig. 1. Inhibition of proliferation of CD4 antigen-specific and non-antigen-specific T cells by Cal g1 and Dox drugs. Negatively sorted CD4 from TCR-HA Tg mice and naı¨ve Balb/c mice were cultured for 3 days with HA-pulsed, irradiated APCs (left panel) or CD3 and CD28 antibodies (right panel), respectively, in the presence of graded amounts of drugs. The 3H-thymidine was added for the last 24 h, and thymidine incorporation was calculated as described. The horizontal dotted lines represent the 50% index of T-cell proliferation, and the vertical dotted lines indicate the corresponding drug concentration. The index of proliferation of CD4 TCR-HA T cells stimulated with HA-pulsed APCs was 82,000 F 950 cpm, and that of naı¨ve CD4 T cells stimulated with CD3/CD28 Abs in the absence of Cal g1 was 60,500 F 1150 cpm.
of time with graded amounts of drugs, and analyzed by electrophoresis in 1.5% agarose gel containing ethidium bromide. For the semi-quantitative RT-PCR analysis of IL-2, IL-4, IL-10 and IFN-g gene transcription, the cDNA was prepared from cells treated with Cal g1 using the Qiagen One Step RT-PCR kit (Qiagen Valencia, CA) and the following specific primers: IL-2
(forward GACACTTGTGCTCCTTGTCA, reverse TCAATTCTGTGGCCTGCTTG), IL-4 (forward TCGGCATTTTGAACGAGGTC, reverse GAAAAGCCCGAAAGAGT CTC), IL-10 (forward 5VATGCAGGACTTTAAGGGTTACTTG, reverse TAGACACCTTG GTCTTGGAGCTTA) and IFN-g (forward GCTCTGAGACAATGAACGCT, reverse
Table 1 Inhibition of cytokine production in T cells by Cal g1 Cytokine production (pg/ml F S.D.)
HA/APC + TCR-HA CD4 T cells + Nil CD3/CD28 Ab + naı¨ve CD4 T cells + Nil Cal g1 (3 ng/ml for 30 min)a + HA/APC + TCR-HA CD4 T cells Cal g1 (3 ng/ml for 30 min)a + CD3/CD28 Ab + naı¨ve CD4 T cells Cal g1 (3 ng/ml for 2 h)a + HA/APC + TCR-HA CD4 T cells Cal g1 (3 ng/ml for 2 h)a + CD3/CD28 Ab + naı¨ve CD4 T cells TCR-HA CD4 T cells + medium alone (background) Naı¨ve CD4 T cells + medium alone (background) a
IL-2
IFN-g
IL-4
IL-10
780 F 3.5 615 F 4.0 280 F 2.0
960 F 2.5 998 F 3.1 48 F 1.8
405 F 1.0 115 F 2.3 10.8 F 1.4
97 F 1.5 12.5 F 2.1 21 F 3.7
141 F 1.7
64 F 2.0
12.2 F 1.0
< 5.5
32 F 3.4
21 F 2.8
10.0 F 2.0
< 5.5
< 5.5
< 5.5
< 5.5
48 F 4.1
37 F 3.1
12 F 2.4
7.5 F 1.8
27 F 3.7
30 F 1.8
7.5 F 2.0
< 5.5
17.5 F 3.1
T cells were pre-treated with Cal g1 for 30 min or 2 h, washed and cultured for 3 days in drug-free medium with the stimulatory HA110120 peptide and irradiated APCs (in the case of TCR-HA CD4 T cells from transgenic mice), or with CD3 plus CD28 stimulatory antibodies (in the case of naı¨ve CD4 T cells from Balb/c mice). Cytokines were measured by ELISA. Shown are the values (pg/ml) from triplicate wells F S.D.
524
S. Thomas et al. / International Immunopharmacology 4 (2004) 521–526
of HA110-120-specific (diabetogenic) CD4 T cells and naı¨ve CD4 T cells. Only 50% of cells proliferated upon continuous incubation with 3 ng/ml of Cal g1, whereas the same effect was obtained with 3000 ng/ ml of Dox. At 100 ng/ml of Cal g1, the HA proliferative response of splenic TCR-HA T cells from Tg mice, and the CD3/CD28-Ab-mediated proliferation of naı¨ve CD4 T cells from Balb/c mice, was abrogated (Fig. 1). The TCR-HA and naı¨ve CD4 thymic T-cell precursors showed similar sensitivity to Cal g1 as the mature T cells (data not shown). The fact that only 50% of T cells ceased proliferation upon a 3-day continuous incubation with 3 ng/ml of Cal g1 suggested that not all cells can uptake the optimal suppressogenic amount of the drug. This raised the question of whether a certain extent of drug degradation may occur in the culture medium. The uptake of Cal g1 by splenic T cells occurred relatively fast, since 30 min pre-incubation with 12.5 ng/ml Cal g1 ceased the HA proliferative response by 80%, and higher doses ceased completeFig. 2. Cal g1 induces fast and severe RNA degradation in TCR-HA T cells. Panel A: the RNA and DNA was extracted from negatively sorted, splenic CD4 TCR-HA T cells, and incubated on ice for various intervals of time and doses, as indicated. The 28s and 18s major RNA bands, and the DNA band were analyzed for the integrity by 1.5% agarose gel electrophoresis. Some 4 Ag/lane of total RNA and DNA material was analyzed. Shown is one of three representative experiments. Panel B: the CD4 TCR-HA T cells (108 cells) were incubated for various intervals of time and doses of Cal g1 as indicated, and the cells were washed and cultured in drug-free medium for another 16 h. The total RNA and DNA was extracted and analyzed as described. Shown is one of two representative experiments carried out with 20 Ag/lane of nucleic material.
AAAGA GATAATCTGGCTGTGC) [8]. The PCR products were separated by electrophoresis in 1.5% agarose gel containing ethidium bromide.
3. Results and discussion 3.1. Suppressogenicity of Cal c1 on T cells is f1000fold higher than that of Dox The index of cell proliferation determined by 3HTdR assay indicated that Cal g1 was f 1000-fold more potent than Dox in suppressing the cell growth
Fig. 3. Cal g1 induces non-selective degradation of interleukins mRNA in TCR-HA T cells. Negatively sorted, splenic CD4 TCRHA T cells were incubated or not for 30 min and 2 h with 5 ng/ml Cal g1, cells were washed and cultured for 16 h with CD3/CD28 stimulatory antibodies. Cells were washed, the total RNA was extracted and cDNA prepared as described. Some 4 Ag/lane of cDNA material was analyzed by RT-PCR using specific primers for IL-2, IL-4, IL-1L and IFN-g. Shown is the amplicon size (base pairs, bp) for each interleukin from one of three representative experiments. Lane 1: cells cultured with medium alone for 16 h in the absence of Cal g1 (control, signal-to-noise background); lane 2: cells stimulated with CD3/CD28 Abs for 16 h in the absence of Cal g1 (control stimulation); lane 3: cells treated for 30 min with Cal g1 (5 ng/ml) and stimulated for 16 h with CD3/CD28 Abs in drug free medium; and lane 4: cells treated for 2 h with Cal g1 (5 ng/ml) and stimulated for 16 h with CD3/CD28 Abs in drug-free medium.
S. Thomas et al. / International Immunopharmacology 4 (2004) 521–526
ly the proliferation. A bystander toxic effect of Cal g1 on APCs has been ruled out in these experiments, since the T cells were first incubated with the drug, and then washed, and incubated with healthy APCs pulsed with HA peptide. These results indicated that Cal g1 toxicity relies mostly on its cellular uptake, and less on its chemical stability. At present, neutralization of Cal g1 toxicity by binding to serum proteins can not be ruled out. This remains to be further investigated. 3.2. Suppressogenicity of Cal c1 on T cells relies mainly on RNA degradation The cell culture supernatants from a 3-day 3H-TdR assay carried out in the presence of toxic doses of Cal g1 showed a lack of IL-2, IFN-g, IL-4 and IL-10 secretion. Similar results were obtained in the case of naı¨ve CD4 T cells from Balb/c mice during stimulation with CD3 and CD28 antibodies (Table 1). This correlated with a profound suppression of the HA proliferative response or CD3/CD28-induced proliferative response, as described (Fig. 1). Analysis of RNA and DNA integrity in TCR-HA T cells suppressed by Cal g1 showed a severe degradation of RNA 30 min after incubation with 5 ng/ml Cal g1, whereas the DNA cleavage occurred 2 h later and at a 10-times higher dose (50 ng/ml) (Fig. 2). The RNA and DNA degradation patterns in naı¨ve CD4 T cells from Balb/c mice were similar to those from TCR-HA T cells (data not shown). It is well known that IL-2 and IL-4 supports the growth of Th1 and Th2 cell subsets, respectively [9]. As the total RNA was rapidly degraded in T cells suppressed by Cal g1, we next analyzed their levels of IL-2, IL-4, IL-10 and IFN-g mRNA expression by semi-quantitative RT-PCR (Fig. 3). Not surprisingly, the mRNA expression for these cytokines was drastically inhibited after 30 min, and completely inhibited after 2 h of incubation with 5 ng/ml of Cal g1. This clearly indicated that a rapid and non-selective RNA degradation was the major cause of Cal g1-induced suppression of cytokine production with subsequent abrogation of cell growth. The mechanism of Cal g1mediated degradation of RNA maybe similar to that described for DNA [2], but far more efficient. This is because RNA is built as a single strand macromolecule, and DNA as a double strand macromolecule.
525
Obviously, a single-strand macromolecule could be easily accessed by the drug. Several reports indicated that RNA degradation is an earlier and more efficient mechanism of cell growth arrest and apoptosis than the DNA degradation [10]. In conclusion, our results demonstrated for the first time that Cal g1 is an extremely potent killing agent for both antigen-specific and non-specific CD4 T cells by a mechanism relying primarily on RNA degradation. The DNA cleavage occurred at higher doses of Cal g1 and after longer time of interaction with the drug. We have previously demonstrated that apoptotic drugs like Doxorubicin can efficiently eliminate diabetogenic T cells in vivo when delivered through a soluble, dimeric peptide-MHC II chimeras (DEF) [6]. Accordingly, one may consider that immunoconjugates of Cal g1 and DEF-like chimeras could provide rational grounds to develop new classes of therapeutic drugs for autoimmune diseases. DEF-Cal g1 conjugates may reduce considerable the therapeutic dose and number of injections, and the systemic toxic effects of the drug in vivo.
Acknowledgements This work was supported by grants from the National Institutes of Health (DK61927 and DK61326) to T.-D.B.
References [1] Love SF, Maiese WM, Rothstein DM. Conditions for protoplasting, regenerating, and transforming the calicheamicin producer, Micromonospora echinospora. Appl Environ Microbiol 1992;58:1376 – 8. [2] Nicolaou KC, Pitsinos N, Theodorakis EA, Saimoto H, Wrasidlo W. Synthetic calicheamicin mimics with novel initiation mechanisms: DNA cleavage, cytotoxicity, and apoptosis. Chem Biol 1994;1:57 – 66. [3] DiJoseph JF, Armellino DC, Boghaert ER, Khandke K, Dougher MM, Sridharan L, et al. Antibody-targeted chemotherapy with CMC-544: a CD22-targeted immunoconjugate of calicheamicin for the treatment of B lymphoid malignancies. Blood 2004;103 (in press). [4] Hamann PR, Hinman LM, Hollander I, Beyer CF, Lindh D, Holcomb R, et al. Gemtuzumab ozogamicin, a potent and selective anti-CD33 antibody-calicheamicin conjugate for treatment of acute myeloid leukemia. Bioconjug Chem 2002; 13:47 – 58.
526
S. Thomas et al. / International Immunopharmacology 4 (2004) 521–526
[5] Stan CA, Radu DL, Casares S, Bona C, Brumeanu TD. Antineoplastic efficacy of Doxorubicin enzymatically assembled on galactose residues of a monoclonal antibody specific for the carcinoembryonic antigen (CEA). Cancer Res 1999;59:115 – 21. [6] Casares S, Bona C, Brumeanu TD. Antigen-specific downregulation of T-cells by doxorubicin delivered by MHC IIpeptide chimeras. Nat Biotechnol 2001;19:142 – 7. [7] Casares S, Hurtado A, von Boehmer H, Sarukhan A, McEvoy R, Brumeanu TD. Downregulation of diabetogenic T-cells by a soluble MHC II-peptide chimera. Nat Immunol 2002; 3:383 – 91.
[8] Zhou HR, Yan D, Pestka JJ. Differential cytokine mRNA expression in mice after oral exposure to the trichothecene vomitoxin (deoxynivalenol):dose response and time course. Toxicol Appl Pharmacol 1997;144:294 – 305. [9] Watowich SS, Wu H, Socolovsky M, Klingmuller U, Constantinescu SN, Lodish HF. Cytokine receptor signal transduction and the control of hematopoietic cell development. Annu Rev Cell Dev Biol 1996;12:91 – 128. [10] Tourriere H, Chebli K, Tazi J. mRNA degradation machines in eukaryotic cells. Biochimie 2002;84:821 – 37.