- Email: [email protected]
DIFFERENTIAL GLYCOSYLATION OF INTERLEUKIN 2, THE MOLECULAR BASIS FOR THE NOD Idd3 TYPE 1 DIABETES GENE?
doi:10.1006/cyto.1999.0609, available online at http://www.idealibrary.com on
SHORT COMMUNICATION
DIFFERENTIAL GLYCOSYLATION OF INTERLEUKIN 2, THE MOLECULAR BASIS FOR THE NOD Idd3 TYPE 1 DIABETES GENE? Patricia L. Podolin,1 Mary Beth Wilusz,2 Rose M. Cubbon,1 Utpal Pajvani,1 Christopher J. Lord,3 John A. Todd,3 Laurence B. Peterson,2 Linda S. Wicker,1 Paul A. Lyons3 The insulin-dependent diabetes (Idd) gene, Idd3, has been localised to a 0.35 cM region of chromosome 3 containing the structural gene for the cytokine interleukin 2 (IL-2). While variation of the N-terminal amino acid sequence of IL-2 has been shown to correlate with Idd3 allelic variation, differences in induction of proliferation by IL-2 allotypes have not been detected. In the current study, we examined the electrophoretic migration of IL-2 allotypes and have found two distinct patterns, consistent with differences in glycosylation, that correlate with diabetes-resistance and susceptibility. These findings strongly suggest that IL-2 variants may be functionally distinct. 2000 Academic Press
The non-obese diabetic (NOD) mouse develops an autoimmune diabetes that is remarkably similar to the human disease. To date, 18 Idd genes have been localised in the NOD mouse. Idd1 and Idd16 map to the MHC, while the remaining 16 genes are non-MHClinked.1–3 Of all the non-MHC genes Idd3 has been localised most precisely, to a 0.35 cM interval in the proximal region of chromosome 3.4–6 The influence of Idd3 on diabetes development was demonstrated in previous studies in which NOD congenic strains homozygous for the B6 or B10-derived Idd3 resistance allele exhibited significantly lower frequencies of diabetes compared to the NOD strain.4 Il2 is a known gene contained within the 0.35 cM Idd3 region.5,6 The IL-2 protein was originally defined as a growth factor that stimulates the differentiation and proliferation of T cells, B cells, NK cells and monocytes/macrophages.7 Surprisingly, in IL-2 or From the Departments of 1Autoimmune Diseases Research and 2 Pharmacology, Merck Research Laboratories, Rahway, New Jersey 07065, USA; 3Cambridge Institute for Medical Research, Department of Medical Genetics, University of Cambridge, Addenbrooke’s Hospital, Hills Road, Cambridge CB2 2XY, UK Correspondence to: Dr Linda Wicker, Merck Research Laboratories, Mail Code RY80W-107, 126 E. Lincoln Ave, Rahway, NJ 07065, USA; E-mail: [email protected] Received 27 January 1999; received in revised form 19 August 1999; accepted for publication 6 September 1999 2000 Academic Press 1043–4666/00/050477+06 $35.00/0 CYTOKINE, Vol. 12, No. 5 (May), 2000: pp 477–482
IL-2-receptor knockout mice these cell subsets were found to develop. However, these knockout strains develop spontaneous autoimmune syndromes illustrating that IL-2 is a non-redundant determinant of T cell tolerance and autoimmunity.8,9 The presence of the Il2 gene within the Idd3 interval, coupled with the pivotal role of this cytokine in the maintenance of selftolerance, makes IL-2 a very strong candidate for Idd3. Consistent with its potential role as Idd3, the diabetesassociated NOD IL-2 molecule and the diabetesresistant B6 molecule differ in their N-terminal sequence.10 However, no difference in the specific activity of these IL-2 molecules as measured by in vitro cell proliferation assays has been detected.11 The goal of the present study was to determine if differences in electrophoretic migration exist between IL-2 molecules derived from the NOD and B6 strains as a consequence of allotypic variation in their N-termini.
RESULTS AND DISCUSSION Correlation between ‘‘homogeneous’’ IL-2 electrophoretic migration and the presence of an Idd3 susceptibility allele To compare the electrophoretic migration of the NOD and B6 IL-2 allotypes, culture supernatants from PMA/ionomycin-stimulated NOD and NOD.B6 Idd3 477
478 / Podolin et al.
CYTOKINE, Vol 12, No. 5 (May, 2000: 477–482)
Figure 1. Differential migration of IL-2 from NOD and NOD.B6 Idd3 mice. IL-2-containing spleen cell supernatants from NOD (Lane 1) or NOD.B6 Idd3 (Lane 3) mice were subjected to SDS-PAGE and Western blot analysis with an anti-IL-2 monoclonal antibody. In lanes 2 and 4, the NOD and NOD.B6 Idd3 supernatants, respectively, were incubated with an anti-IL-2 antibody followed by adsorption with anti-IgG-sepharose to demonstrate that all bands were IL-2.
spleen cells were analysed by SDS-PAGE and Western blot analysis (Figs 1 and 2). NOD IL-2 migrated principally as a 21–22 kDa species, with a smaller, minor species, apparent as well (Fig. 1, lane 1 and Fig. 2, lane 2). In contrast, B6 IL-2 (from the NOD.B6 Idd3 strain) demonstrated a more heterogeneous pattern of migration with several major species ranging between 17 and 19 kDa (Fig. 1, lane 3 and Fig. 2, lane 3). Our previous studies have shown that the NOD strain is homozygous for an Idd3 susceptibility allele whereas the B6 strain is homozygous for an Idd3 resistance allele.5,6 Thus, a ‘‘homogeneous’’ IL-2 electrophoretic pattern (as for the NOD strain IL-2 allotype) correlates with the presence of an Idd3 susceptibility genotype and the ‘‘heterogeneous’’ electrophoretic pattern (as for B6 IL-2) correlates with the Idd3 resistance genotype. In order to establish further the relationship between Idd3 genotype and IL-2 migration pattern, we determined the Idd3 genotype of the 129 mouse and compared it with the electrophoretic migration of IL-2 derived from this allele. The sequence of the 129 IL-2 molecule was determined using a strategy described previously6 and found to differ from NOD IL-2 by a single amino acid residue, a proline substitution at
position four (Table 1). To assess the Idd3 status of the 129 strain, we developed a NOD.129 Idd3 congenic strain that has a 129-derived introgressed segment on chromosome 3 between, and including, the microsatellite markers D3Mit93 (13.8 cM) and D3Mit65 (23.3 cM). This 9.5 cM interval encompasses Idd3 and the Il2 gene but does not include either Idd10, Idd17, or Idd18.12,13 The frequency of diabetes in females of the NOD.129 Idd3 congenic strain at 7 months of age was 75.3% (60/81) compared to 77.8% (63/81) in NOD (P>0.05) and 20.3% (12/59) in the NOD.B6 Idd3 strain (P<0.0001) (Fig. 3). Thus 129, like NOD, has a diabetes susceptibility allele at Idd3. IL-2-containing T cell supernatants from NOD.129 Idd3 mice were then compared to those obtained from NOD and NOD.B6 Idd3 mice (Figs 2 and 4). 129 IL-2 has a homogeneous electrophoretic pattern supporting the correlation between Idd3 genotype and IL-2 electrophoretic migration.
Correlation between IL-2 electrophoretic migration and IL-2 residue 6 To establish any correlation between IL-2 N-terminal amino acid sequence and IL-2 electrophoretic mobility, SDS-PAGE and Western blot analysis were performed on culture supernatants from
Differential glycosylation of IL-2 and diabetes / 479
Figure 2.
Correlation between IL-2 electrophoretic migration and IL-2 residue 6.
Supernatants from activated splenocytes obtained from NOD, NOD.B6 Idd3, NOD.129 Idd3, Spretus, and Czech mice (Lanes 2–6, respectively) were assessed for IL-2 by SDS-PAGE and Western blot analysis. Lane 1 contains recombinant IL-2 (B6 Il2 gene) produced in T. ni cells.
TABLE 1.
Comparison of IL-2 electrophoretic migration patterns and N-terminal amino acid sequences
Strain NOD, A/J B10.H2g7 Idd3nod
Electrophoretic pattern Homo*
N-terminal amino acid sequence Ala Pro Thr
Ser Ser
Pro
Thr Ser Ser
Pro
Thr
Ser Ser Ser Thr Ala Glu Ala
(Gln)8
NOD.129 Idd3
Homo
Ala Pro Thr
Pro Ser
Pro Thr Ser Ser
Pro Thr
Ser Ser Ser Thr Ala Glu Ala
(Gln)8
SPRET
Homo
Ala Pro Thr
Ser
Ser
Pro Thr Ser Ser
Ser
Thr
Ser Ser Ser Thr Ala Glu Ala
(Gln)5
CZECH
Homo
Ala Pro Thr
Ser
Ser
Pro Thr Ser Ser
Ser
Thr
—
—
—
—
Ala Glu Ala
(Gln)21
B6, BALB/c NOD.B6 Idd3 B10.H2g7
Hetero
Ala Pro Thr
Ser
Ser
Ser
Ser
Thr
—
—
—
—
Ala Glu Ala
(Gln)12
Thr Ser Ser
*Homo and hetero indicate a homogeneous and heterogeneous electrophoretic migration pattern, respectively.
additional mouse strains. The IL-2 N-terminal amino acid sequence of each strain predicted from the Il2 exon 1 DNA sequence is shown in Table 1. Like IL-2 secreted from NOD.B6 Idd3 cells, IL-2 from B6, BALB/c, and B10.H2g7 cells (Fig. 4) exhibits a ‘‘heterogeneous’’ pattern of electrophoretic mobility. In contrast, the electrophoretic mobility of secreted IL-2 from A/J, SPRET, CZECH, and B10.H2g7 Idd3nod cells is ‘‘homogeneous’’ (Figs 2 and 4). The amino acid sequence of the five IL-2 variants studied differ in the presence of a serine versus a proline at residues 4, 6, and 10 of the mature protein,
the number of proline/serine-threonine-serine-serine repeats (2.5 or 3.5 repeats), and the number of glutamine residues constituting the polyglutamine sequence (Table 1). The only sequence variant that correlates with electrophoretic mobility is the presence of a serine or proline residue at position six of the mature IL-2 protein. Serine at position six of IL-2 correlates with a heterogeneous electrophoretic mobility (B6 and BALB/c), whereas proline at position six is found only in strains with a homogeneous electrophoretic mobility (NOD, A/J, 129, SPRET and CZECH).
480 / Podolin et al.
CYTOKINE, Vol 12, No. 5 (May, 2000: 477–482)
Cumulative frequency of diabetes (%)
100 90 80 70 60 50 40 30 20 10 0 60
80
100
120 140 160 Age (days)
180
200
220
Figure 3. The 129-derived Idd3 gene does not protect NOD mice from diabetes. Female NOD (n=73, — —), NOD.129 Idd3 (n=81, — —), and NOD.B6 Idd3 (n=59, — —) mice were monitored for the development of diabetes as reported previously.4 Fisher’s Exact Test was used to compare the 7 month cumulative diabetes frequencies.
The multiple apparent molecular mass species of IL-2 may represent different glycosylation forms One likely explanation for the homogeneous and heterogeneous patterns is that there are differences in glycosylation between the IL-2 allotypes. The amino acid sequences of the N-terminus of human and mouse IL-2 are highly conserved and the mature human IL-2 protein has been shown to be O-glycosylated at the
Figure 4.
position three threonine residue.14 Human and B6 IL-2, which share a serine at position six, exhibit similar electrophoretic heterogeneity;14,15 this heterogeneity in human IL-2 has been shown to be due to variable glycosylation at position three.16 In addition, purification of mouse IL-2 has shown that it is comprised of several species differing in molecular mass and isoelectric point which is consistent with differences in glycosylation.15 The electrophoretic mobility of T cell-derived IL-2 encoded by the B6 and NOD Il2 alleles was compared with commercially available recombinant B6 IL-2 proteins. The recombinant IL-2 is produced by cells that are capable of glycosylation (baculovirus-infected Trichoplusia ni cells) or by glycosylation-incompetent cells (Escherichia coli). As seen in Fig. 4, E. coli-derived B6 IL-2 migrates as a single species, the apparent molecular mass of which is consistent with the predicted mass of 17.1 kDa. A portion of the B6 IL-2 derived from T. ni cells or from mouse spleen cells migrates closely with the unglycosylated form of IL-2 derived from E. coli. In addition, IL-2 secreted from both eukaryotic cell types has higher molecular mass species, suggesting that these are glycosylated forms of the protein. Interestingly, even though the B6 IL-2 was expressed in evolutionarily distant eukaryotic cells the patterns were similar (Figs 2 and 4). In contrast, only very small amounts of unglycosylated IL-2 are secreted by NOD spleen cells, with most of the secreted IL-2 represented by a single, presumably glycosylated
The alternative patterns of IL-2 migration may represent different glycosylation forms.
IL-2-containing supernatants (upper panel) and cell lysates (lower panel) from activated splenocytes obtained from the indicated mouse strains were analysed by SDS-PAGE and Western blot analysis. In each panel, 100 and 500 pg of recombinant B6 IL-2 purified from E. coli (glycosylation incompetent) or Baculovirus-infected T. ni (glycosylation competent) are shown for comparison.
Differential glycosylation of IL-2 and diabetes / 481
Concluding remarks
350 A 300
CPM (×10–3)
250 200 150 100 50 0
0.01
0.1
1
IL-2 (ng/ml) 300 B
CPM (×10–3)
250 200 150 100 50 0
0.01
0.1 IL-2 (ng/ml)
We demonstrate for the first time that there are two forms of mouse IL-2 that are elecrophoretically distinct and that these forms are determined by amino acid variation at position six of the mature IL-2 protein. The presence of serine at position six results in the secretion of IL-2 molecules that migrate in a heterogeneous pattern consistent with the presence of both unglycosylated and multiple glycosylated species. In contrast, the presence of proline at position six results in nearly all of the secreted IL-2 being found as a single glycosylated species. Using congenic mouse strains we have found that diabetes resistance encoded by Idd3 correlates with the production of unglycosylated and multiple glycosylated IL-2 molecules, whereas diabetes susceptibility is associated with primarily a single glycosylated species. If this variation is the molecular basis of Idd3-mediated diabetes resistance versus susceptibility, the mechanism is not via a difference in the ability of the two forms to induce IL-2-dependent proliferation, at least in transformed cells. Other potential mechanisms that can be considered are effects on in vivo half-life of the glycosylated and unglycosylated forms, differential binding to extracellular matrix components, and differential activation of non-proliferative signal transduction pathways such as those leading to activation-induced cell death.
1
Figure 5. IL-2 from NOD ( ), NOD.B6 Idd3 ( ) and NOD.129 Idd3 () splenocytes induce equivalent proliferation of the IL-2dependent cell lines CTLL-2 (a) and HT-2 (b). IL-2 dependence of proliferation induced by the activated T cell supernatants was confirmed by the observation that complete inhibition of proliferation was achieved by the addition of a neutralising anti-IL-2 monoclonal antibody (data not shown).
species that migrates more slowly than any of the species encoded by the B6 Il2 gene. Consistent with this interpretation is our observation that intracellular IL-2 derived from all Il2 alleles examined has a species of IL-2 which co-migrates with unglycosylated IL-2 (Fig. 4).
Proliferative activity of IL-2 allotypes Given the striking differences in the electrophoretic mobility of IL-2 allotypes, IL-2 obtained from NOD, NOD.B6 Idd3 and NOD.129 Idd3 spleen cells activated with ionomycin and PMA was tested for its ability to stimulate proliferation of the IL-2-dependent cell lines, CTLL-2 and HT-2. All three sources of IL-2 produce equivalent levels of IL-2-dependent proliferation (Fig. 5).
MATERIALS AND METHODS Animals NOD and 129 mice were obtained from Taconic Farms (Germantown, NY, USA). The derivation of the B10.H2g7 and the B10.H2g7 Idd3nod strains17 as well as the NOD.B6 Idd3 strain6 have been described previously. The NOD.129 Idd3 strain was utilised at the N6F2-4 generations and its development was essentially as described for the NOD.B6 Idd3 strain. All other mice were obtained from The Jackson Laboratory (Bar Harbor, ME, USA).
Generation of secreted and intracellular IL-2 Murine spleen cells (8105) were stimulated for 24 h with 10 ng/ml of phorbol 2-myristate 13-acetate (PMA, Sigma Chemical Co., St. Louis, MO, USA) and 400 ng/ml of ionomycin (Calbiochem, San Diego, CA, USA) using culture conditions described previously.18 Intracellular IL-2 was obtained from whole cell lysates following incubation with PMA and ionomycin for 4 h. Cell pellets were resuspended in lysis buffer [50 mM Tris (pH 8.0), 150 mM NaCl, 2 mM EDTA, 0.2 mM PMSF, 1 g/ml leupeptin, 1 g/ml aprotinin, 10 g/ml chymostatin, 1 mM NaF, 1 mM sodium orthovanadate plus 1% Nonidet P-40] at 108 cells per ml and incubated on ice for 30 min. Cellular debris was removed by centrifugation.
482 / Podolin et al.
SDS-PAGE and Western Blot analysis SDS-PAGE was performed on recombinant mouse IL-2 samples (the B6 allele expressed in baculovirus-infected Trichoplusia ni cells, PharMingen19 and the B6 allele expressed in E. coli, R&D Systems, Minneapolis, MN, USA) and IL-2-containing culture supernatants and cell lysates according to the procedure of Laemmli.20 Except as noted in Figure 4, 50–200 pg of IL-2 was analysed per sample. Proteins were transferred onto nitrocellulose membranes using standard procedures.
IL-2 quantitation IL-2 protein levels were determined as per Chen et al.18 IL-2 bioactivity was assessed on CTLL-2 and HT-2 cells (American Type Culture Collection, Rockville, MD, USA) as previously described.11 To confirm that proliferation was caused only by IL-2 within the culture supernatants, a neutralising anti-IL-2 antibody (clone S4B6, Pharmingen) was added at 5 g/ml.
REFERENCES 1. Wicker LS, Todd JA, Peterson LB (1995) Genetic control of autoimmune diabetes in the NOD mouse. Ann Rev Immunol 13:179–200. 2. Leiter EH (1998) Genetics and immunogenetics of NOD mice and related strains. In Leiter EH, Atkinson MA (eds) NOD mice and related strains: Research Applications in Diabetes, AIDS, Cancer and Other Diseases. Landes Bioscience Publishers, Austin, pp 37–69. 3. Lyons PA, Wicker LS (1999) Localising polygenes in the NOD mouse model of type 1 diabetes. In Theofilopoulos A (ed) Genetics in Autoimmunity. Karger, Basel, pp 208–225. 4. Wicker LS, Todd JA, Prins J-B, Podolin PL, Renjilian RJ, Peterson LB (1994) Resistance alleles at two non-major histocompatibility complex-linked insulin-dependent diabetes loci on chromosome 3, Idd3 and Idd10, protect nonobese diabetic mice from diabetes. J Exp Med 180:1705–1713. 5. Lord CJ, Bohlander SK, Hopes EA, Montague CT, Hill NJ, Prins J-B, Peterson LB, Wicker LS, Todd JA, Denny P (1995) Mapping of the diabetes polygene Idd3 on mouse chromosome 3 using novel congenic strains. Mamm Genome 6:563–570. 6. Denny P, Lord CJ, Hill NJ, Goy JV, Levy ER, Podolin PL, Peterson LB, Wicker LS, Todd JA, Lyons PA (1997) Mapping
CYTOKINE, Vol 12, No. 5 (May, 2000: 477–482) of the IDDM locus Idd3 to a 0.35 cM interval containing the Interleukin-2 gene. Diabetes 46:695–700. 7. Smith KA (1992) Interleukin-2. Curr Opin Immunol 4: 271–276. 8. Hunig T, Schimpl A (1998) The IL-2 deficiency syndrome. A lethal disease caused by abnormal lymphocyte survival. In Durum SK, Muregge K (eds) Cytokine Knockouts. Humana Press Inc, Tolowa, pp 1–19. 9. Van Parijs L, Abbas AK (1998) Homeostasis and selftolerance in the immune system: Turning lymphocytes off. Science 280:243–248. 10. Ghosh S, Palmer SM, Rodrigues NR, Cordell HJ, Hearne CM, Cornall R, Prins J-B, McShane P, Lathrop GM, Peterson LB, Wicker LS, Todd JA (1993) Polygenic control of autoimmune diabetes in nonobese diabetic mice. Nature Genet 4:404–409. 11. Chesnut K, She J-X, Cheng I, Muralidharan K, Wakeland EK (1993) Characterizations of candidate genes for IDD susceptibility from the diabetes-prone NOD mouse strain. Mamm Genome 4:549–554. 12. Podolin PL, Denny P, Lord CJ, Hill NJ, Todd JA, Peterson LB, Wicker LS, Lyons PA (1997) Congenic mapping of the insulin dependent diabetes (Idd) gene, Idd10, localizes two genes mediating the Idd10 effect, and eliminates the candidate Fcgr1. J Immunol 159:1835–1843. 13. Podolin PL, Denny P, Armitage N, Lord CJ, Hill NJ, Levy ER, Peterson LB, Todd JA, Wicker LS, Lyons PA (1998) Localization of two insulin-dependent diabetes (Idd) genes to the Idd10 region on mouse chromosome 3. Mamm Genome 9:283–286. 14. Conradt HS, Geyer R, Hoppe J, Grotjahn L, Plessing A, Mohr H (1985) Structures of the major carbohydrates of natural human interleukin-2. Eur J Biochem 153:255–261. 15. Gillis S, Mochizuki DY, Conlon PJ, Hefeneider SH, Ramthun CA, Gillis AE, Frank MB, Henney CS, Watson JD (1982) Molecular characterization of interleukin 2. Immunol Rev 63: 167–209. 16. Robb RJ, Smith KA (1981) Heterogeneity of human T-cell growth factor(s) due to variable glycosylation. Mol Immunol 18:1087–1094. 17. Colomb E, Savino W, Wicker L, Peterson L, Dardenne M, Carnaud C (1996) Genetic control of giant perivascular space formation in the thymus of NOD mice. Diabetes 45:1535–1540. 18. Chen SL, Whiteley PJ, Freed DC, Rothbard JB, Peterson LB, Wicker LS (1994) Responses of NOD congenic mice to a glutamic acid decarboxylase-derived peptide. J Autoimmun 7: 635–641. 19. Kashima N, Nishi-Takaoka C, Fujita T, Taki S, Yamada G, Hamuro J, Taniguchi T (1985) Unique structure of murine interleukin-2 as deduced from cloned cDNAs. Nature 313:402–404. 20. Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of the bacteriophage T4. Nature 227: 680–685.
SHORT COMMUNICATION
DIFFERENTIAL GLYCOSYLATION OF INTERLEUKIN 2, THE MOLECULAR BASIS FOR THE NOD Idd3 TYPE 1 DIABETES GENE? Patricia L. Podolin,1 Mary Beth Wilusz,2 Rose M. Cubbon,1 Utpal Pajvani,1 Christopher J. Lord,3 John A. Todd,3 Laurence B. Peterson,2 Linda S. Wicker,1 Paul A. Lyons3 The insulin-dependent diabetes (Idd) gene, Idd3, has been localised to a 0.35 cM region of chromosome 3 containing the structural gene for the cytokine interleukin 2 (IL-2). While variation of the N-terminal amino acid sequence of IL-2 has been shown to correlate with Idd3 allelic variation, differences in induction of proliferation by IL-2 allotypes have not been detected. In the current study, we examined the electrophoretic migration of IL-2 allotypes and have found two distinct patterns, consistent with differences in glycosylation, that correlate with diabetes-resistance and susceptibility. These findings strongly suggest that IL-2 variants may be functionally distinct. 2000 Academic Press
The non-obese diabetic (NOD) mouse develops an autoimmune diabetes that is remarkably similar to the human disease. To date, 18 Idd genes have been localised in the NOD mouse. Idd1 and Idd16 map to the MHC, while the remaining 16 genes are non-MHClinked.1–3 Of all the non-MHC genes Idd3 has been localised most precisely, to a 0.35 cM interval in the proximal region of chromosome 3.4–6 The influence of Idd3 on diabetes development was demonstrated in previous studies in which NOD congenic strains homozygous for the B6 or B10-derived Idd3 resistance allele exhibited significantly lower frequencies of diabetes compared to the NOD strain.4 Il2 is a known gene contained within the 0.35 cM Idd3 region.5,6 The IL-2 protein was originally defined as a growth factor that stimulates the differentiation and proliferation of T cells, B cells, NK cells and monocytes/macrophages.7 Surprisingly, in IL-2 or From the Departments of 1Autoimmune Diseases Research and 2 Pharmacology, Merck Research Laboratories, Rahway, New Jersey 07065, USA; 3Cambridge Institute for Medical Research, Department of Medical Genetics, University of Cambridge, Addenbrooke’s Hospital, Hills Road, Cambridge CB2 2XY, UK Correspondence to: Dr Linda Wicker, Merck Research Laboratories, Mail Code RY80W-107, 126 E. Lincoln Ave, Rahway, NJ 07065, USA; E-mail: [email protected] Received 27 January 1999; received in revised form 19 August 1999; accepted for publication 6 September 1999 2000 Academic Press 1043–4666/00/050477+06 $35.00/0 CYTOKINE, Vol. 12, No. 5 (May), 2000: pp 477–482
IL-2-receptor knockout mice these cell subsets were found to develop. However, these knockout strains develop spontaneous autoimmune syndromes illustrating that IL-2 is a non-redundant determinant of T cell tolerance and autoimmunity.8,9 The presence of the Il2 gene within the Idd3 interval, coupled with the pivotal role of this cytokine in the maintenance of selftolerance, makes IL-2 a very strong candidate for Idd3. Consistent with its potential role as Idd3, the diabetesassociated NOD IL-2 molecule and the diabetesresistant B6 molecule differ in their N-terminal sequence.10 However, no difference in the specific activity of these IL-2 molecules as measured by in vitro cell proliferation assays has been detected.11 The goal of the present study was to determine if differences in electrophoretic migration exist between IL-2 molecules derived from the NOD and B6 strains as a consequence of allotypic variation in their N-termini.
RESULTS AND DISCUSSION Correlation between ‘‘homogeneous’’ IL-2 electrophoretic migration and the presence of an Idd3 susceptibility allele To compare the electrophoretic migration of the NOD and B6 IL-2 allotypes, culture supernatants from PMA/ionomycin-stimulated NOD and NOD.B6 Idd3 477
478 / Podolin et al.
CYTOKINE, Vol 12, No. 5 (May, 2000: 477–482)
Figure 1. Differential migration of IL-2 from NOD and NOD.B6 Idd3 mice. IL-2-containing spleen cell supernatants from NOD (Lane 1) or NOD.B6 Idd3 (Lane 3) mice were subjected to SDS-PAGE and Western blot analysis with an anti-IL-2 monoclonal antibody. In lanes 2 and 4, the NOD and NOD.B6 Idd3 supernatants, respectively, were incubated with an anti-IL-2 antibody followed by adsorption with anti-IgG-sepharose to demonstrate that all bands were IL-2.
spleen cells were analysed by SDS-PAGE and Western blot analysis (Figs 1 and 2). NOD IL-2 migrated principally as a 21–22 kDa species, with a smaller, minor species, apparent as well (Fig. 1, lane 1 and Fig. 2, lane 2). In contrast, B6 IL-2 (from the NOD.B6 Idd3 strain) demonstrated a more heterogeneous pattern of migration with several major species ranging between 17 and 19 kDa (Fig. 1, lane 3 and Fig. 2, lane 3). Our previous studies have shown that the NOD strain is homozygous for an Idd3 susceptibility allele whereas the B6 strain is homozygous for an Idd3 resistance allele.5,6 Thus, a ‘‘homogeneous’’ IL-2 electrophoretic pattern (as for the NOD strain IL-2 allotype) correlates with the presence of an Idd3 susceptibility genotype and the ‘‘heterogeneous’’ electrophoretic pattern (as for B6 IL-2) correlates with the Idd3 resistance genotype. In order to establish further the relationship between Idd3 genotype and IL-2 migration pattern, we determined the Idd3 genotype of the 129 mouse and compared it with the electrophoretic migration of IL-2 derived from this allele. The sequence of the 129 IL-2 molecule was determined using a strategy described previously6 and found to differ from NOD IL-2 by a single amino acid residue, a proline substitution at
position four (Table 1). To assess the Idd3 status of the 129 strain, we developed a NOD.129 Idd3 congenic strain that has a 129-derived introgressed segment on chromosome 3 between, and including, the microsatellite markers D3Mit93 (13.8 cM) and D3Mit65 (23.3 cM). This 9.5 cM interval encompasses Idd3 and the Il2 gene but does not include either Idd10, Idd17, or Idd18.12,13 The frequency of diabetes in females of the NOD.129 Idd3 congenic strain at 7 months of age was 75.3% (60/81) compared to 77.8% (63/81) in NOD (P>0.05) and 20.3% (12/59) in the NOD.B6 Idd3 strain (P<0.0001) (Fig. 3). Thus 129, like NOD, has a diabetes susceptibility allele at Idd3. IL-2-containing T cell supernatants from NOD.129 Idd3 mice were then compared to those obtained from NOD and NOD.B6 Idd3 mice (Figs 2 and 4). 129 IL-2 has a homogeneous electrophoretic pattern supporting the correlation between Idd3 genotype and IL-2 electrophoretic migration.
Correlation between IL-2 electrophoretic migration and IL-2 residue 6 To establish any correlation between IL-2 N-terminal amino acid sequence and IL-2 electrophoretic mobility, SDS-PAGE and Western blot analysis were performed on culture supernatants from
Differential glycosylation of IL-2 and diabetes / 479
Figure 2.
Correlation between IL-2 electrophoretic migration and IL-2 residue 6.
Supernatants from activated splenocytes obtained from NOD, NOD.B6 Idd3, NOD.129 Idd3, Spretus, and Czech mice (Lanes 2–6, respectively) were assessed for IL-2 by SDS-PAGE and Western blot analysis. Lane 1 contains recombinant IL-2 (B6 Il2 gene) produced in T. ni cells.
TABLE 1.
Comparison of IL-2 electrophoretic migration patterns and N-terminal amino acid sequences
Strain NOD, A/J B10.H2g7 Idd3nod
Electrophoretic pattern Homo*
N-terminal amino acid sequence Ala Pro Thr
Ser Ser
Pro
Thr Ser Ser
Pro
Thr
Ser Ser Ser Thr Ala Glu Ala
(Gln)8
NOD.129 Idd3
Homo
Ala Pro Thr
Pro Ser
Pro Thr Ser Ser
Pro Thr
Ser Ser Ser Thr Ala Glu Ala
(Gln)8
SPRET
Homo
Ala Pro Thr
Ser
Ser
Pro Thr Ser Ser
Ser
Thr
Ser Ser Ser Thr Ala Glu Ala
(Gln)5
CZECH
Homo
Ala Pro Thr
Ser
Ser
Pro Thr Ser Ser
Ser
Thr
—
—
—
—
Ala Glu Ala
(Gln)21
B6, BALB/c NOD.B6 Idd3 B10.H2g7
Hetero
Ala Pro Thr
Ser
Ser
Ser
Ser
Thr
—
—
—
—
Ala Glu Ala
(Gln)12
Thr Ser Ser
*Homo and hetero indicate a homogeneous and heterogeneous electrophoretic migration pattern, respectively.
additional mouse strains. The IL-2 N-terminal amino acid sequence of each strain predicted from the Il2 exon 1 DNA sequence is shown in Table 1. Like IL-2 secreted from NOD.B6 Idd3 cells, IL-2 from B6, BALB/c, and B10.H2g7 cells (Fig. 4) exhibits a ‘‘heterogeneous’’ pattern of electrophoretic mobility. In contrast, the electrophoretic mobility of secreted IL-2 from A/J, SPRET, CZECH, and B10.H2g7 Idd3nod cells is ‘‘homogeneous’’ (Figs 2 and 4). The amino acid sequence of the five IL-2 variants studied differ in the presence of a serine versus a proline at residues 4, 6, and 10 of the mature protein,
the number of proline/serine-threonine-serine-serine repeats (2.5 or 3.5 repeats), and the number of glutamine residues constituting the polyglutamine sequence (Table 1). The only sequence variant that correlates with electrophoretic mobility is the presence of a serine or proline residue at position six of the mature IL-2 protein. Serine at position six of IL-2 correlates with a heterogeneous electrophoretic mobility (B6 and BALB/c), whereas proline at position six is found only in strains with a homogeneous electrophoretic mobility (NOD, A/J, 129, SPRET and CZECH).
480 / Podolin et al.
CYTOKINE, Vol 12, No. 5 (May, 2000: 477–482)
Cumulative frequency of diabetes (%)
100 90 80 70 60 50 40 30 20 10 0 60
80
100
120 140 160 Age (days)
180
200
220
Figure 3. The 129-derived Idd3 gene does not protect NOD mice from diabetes. Female NOD (n=73, — —), NOD.129 Idd3 (n=81, — —), and NOD.B6 Idd3 (n=59, — —) mice were monitored for the development of diabetes as reported previously.4 Fisher’s Exact Test was used to compare the 7 month cumulative diabetes frequencies.
The multiple apparent molecular mass species of IL-2 may represent different glycosylation forms One likely explanation for the homogeneous and heterogeneous patterns is that there are differences in glycosylation between the IL-2 allotypes. The amino acid sequences of the N-terminus of human and mouse IL-2 are highly conserved and the mature human IL-2 protein has been shown to be O-glycosylated at the
Figure 4.
position three threonine residue.14 Human and B6 IL-2, which share a serine at position six, exhibit similar electrophoretic heterogeneity;14,15 this heterogeneity in human IL-2 has been shown to be due to variable glycosylation at position three.16 In addition, purification of mouse IL-2 has shown that it is comprised of several species differing in molecular mass and isoelectric point which is consistent with differences in glycosylation.15 The electrophoretic mobility of T cell-derived IL-2 encoded by the B6 and NOD Il2 alleles was compared with commercially available recombinant B6 IL-2 proteins. The recombinant IL-2 is produced by cells that are capable of glycosylation (baculovirus-infected Trichoplusia ni cells) or by glycosylation-incompetent cells (Escherichia coli). As seen in Fig. 4, E. coli-derived B6 IL-2 migrates as a single species, the apparent molecular mass of which is consistent with the predicted mass of 17.1 kDa. A portion of the B6 IL-2 derived from T. ni cells or from mouse spleen cells migrates closely with the unglycosylated form of IL-2 derived from E. coli. In addition, IL-2 secreted from both eukaryotic cell types has higher molecular mass species, suggesting that these are glycosylated forms of the protein. Interestingly, even though the B6 IL-2 was expressed in evolutionarily distant eukaryotic cells the patterns were similar (Figs 2 and 4). In contrast, only very small amounts of unglycosylated IL-2 are secreted by NOD spleen cells, with most of the secreted IL-2 represented by a single, presumably glycosylated
The alternative patterns of IL-2 migration may represent different glycosylation forms.
IL-2-containing supernatants (upper panel) and cell lysates (lower panel) from activated splenocytes obtained from the indicated mouse strains were analysed by SDS-PAGE and Western blot analysis. In each panel, 100 and 500 pg of recombinant B6 IL-2 purified from E. coli (glycosylation incompetent) or Baculovirus-infected T. ni (glycosylation competent) are shown for comparison.
Differential glycosylation of IL-2 and diabetes / 481
Concluding remarks
350 A 300
CPM (×10–3)
250 200 150 100 50 0
0.01
0.1
1
IL-2 (ng/ml) 300 B
CPM (×10–3)
250 200 150 100 50 0
0.01
0.1 IL-2 (ng/ml)
We demonstrate for the first time that there are two forms of mouse IL-2 that are elecrophoretically distinct and that these forms are determined by amino acid variation at position six of the mature IL-2 protein. The presence of serine at position six results in the secretion of IL-2 molecules that migrate in a heterogeneous pattern consistent with the presence of both unglycosylated and multiple glycosylated species. In contrast, the presence of proline at position six results in nearly all of the secreted IL-2 being found as a single glycosylated species. Using congenic mouse strains we have found that diabetes resistance encoded by Idd3 correlates with the production of unglycosylated and multiple glycosylated IL-2 molecules, whereas diabetes susceptibility is associated with primarily a single glycosylated species. If this variation is the molecular basis of Idd3-mediated diabetes resistance versus susceptibility, the mechanism is not via a difference in the ability of the two forms to induce IL-2-dependent proliferation, at least in transformed cells. Other potential mechanisms that can be considered are effects on in vivo half-life of the glycosylated and unglycosylated forms, differential binding to extracellular matrix components, and differential activation of non-proliferative signal transduction pathways such as those leading to activation-induced cell death.
1
Figure 5. IL-2 from NOD ( ), NOD.B6 Idd3 ( ) and NOD.129 Idd3 () splenocytes induce equivalent proliferation of the IL-2dependent cell lines CTLL-2 (a) and HT-2 (b). IL-2 dependence of proliferation induced by the activated T cell supernatants was confirmed by the observation that complete inhibition of proliferation was achieved by the addition of a neutralising anti-IL-2 monoclonal antibody (data not shown).
species that migrates more slowly than any of the species encoded by the B6 Il2 gene. Consistent with this interpretation is our observation that intracellular IL-2 derived from all Il2 alleles examined has a species of IL-2 which co-migrates with unglycosylated IL-2 (Fig. 4).
Proliferative activity of IL-2 allotypes Given the striking differences in the electrophoretic mobility of IL-2 allotypes, IL-2 obtained from NOD, NOD.B6 Idd3 and NOD.129 Idd3 spleen cells activated with ionomycin and PMA was tested for its ability to stimulate proliferation of the IL-2-dependent cell lines, CTLL-2 and HT-2. All three sources of IL-2 produce equivalent levels of IL-2-dependent proliferation (Fig. 5).
MATERIALS AND METHODS Animals NOD and 129 mice were obtained from Taconic Farms (Germantown, NY, USA). The derivation of the B10.H2g7 and the B10.H2g7 Idd3nod strains17 as well as the NOD.B6 Idd3 strain6 have been described previously. The NOD.129 Idd3 strain was utilised at the N6F2-4 generations and its development was essentially as described for the NOD.B6 Idd3 strain. All other mice were obtained from The Jackson Laboratory (Bar Harbor, ME, USA).
Generation of secreted and intracellular IL-2 Murine spleen cells (8105) were stimulated for 24 h with 10 ng/ml of phorbol 2-myristate 13-acetate (PMA, Sigma Chemical Co., St. Louis, MO, USA) and 400 ng/ml of ionomycin (Calbiochem, San Diego, CA, USA) using culture conditions described previously.18 Intracellular IL-2 was obtained from whole cell lysates following incubation with PMA and ionomycin for 4 h. Cell pellets were resuspended in lysis buffer [50 mM Tris (pH 8.0), 150 mM NaCl, 2 mM EDTA, 0.2 mM PMSF, 1 g/ml leupeptin, 1 g/ml aprotinin, 10 g/ml chymostatin, 1 mM NaF, 1 mM sodium orthovanadate plus 1% Nonidet P-40] at 108 cells per ml and incubated on ice for 30 min. Cellular debris was removed by centrifugation.
482 / Podolin et al.
SDS-PAGE and Western Blot analysis SDS-PAGE was performed on recombinant mouse IL-2 samples (the B6 allele expressed in baculovirus-infected Trichoplusia ni cells, PharMingen19 and the B6 allele expressed in E. coli, R&D Systems, Minneapolis, MN, USA) and IL-2-containing culture supernatants and cell lysates according to the procedure of Laemmli.20 Except as noted in Figure 4, 50–200 pg of IL-2 was analysed per sample. Proteins were transferred onto nitrocellulose membranes using standard procedures.
IL-2 quantitation IL-2 protein levels were determined as per Chen et al.18 IL-2 bioactivity was assessed on CTLL-2 and HT-2 cells (American Type Culture Collection, Rockville, MD, USA) as previously described.11 To confirm that proliferation was caused only by IL-2 within the culture supernatants, a neutralising anti-IL-2 antibody (clone S4B6, Pharmingen) was added at 5 g/ml.
REFERENCES 1. Wicker LS, Todd JA, Peterson LB (1995) Genetic control of autoimmune diabetes in the NOD mouse. Ann Rev Immunol 13:179–200. 2. Leiter EH (1998) Genetics and immunogenetics of NOD mice and related strains. In Leiter EH, Atkinson MA (eds) NOD mice and related strains: Research Applications in Diabetes, AIDS, Cancer and Other Diseases. Landes Bioscience Publishers, Austin, pp 37–69. 3. Lyons PA, Wicker LS (1999) Localising polygenes in the NOD mouse model of type 1 diabetes. In Theofilopoulos A (ed) Genetics in Autoimmunity. Karger, Basel, pp 208–225. 4. Wicker LS, Todd JA, Prins J-B, Podolin PL, Renjilian RJ, Peterson LB (1994) Resistance alleles at two non-major histocompatibility complex-linked insulin-dependent diabetes loci on chromosome 3, Idd3 and Idd10, protect nonobese diabetic mice from diabetes. J Exp Med 180:1705–1713. 5. Lord CJ, Bohlander SK, Hopes EA, Montague CT, Hill NJ, Prins J-B, Peterson LB, Wicker LS, Todd JA, Denny P (1995) Mapping of the diabetes polygene Idd3 on mouse chromosome 3 using novel congenic strains. Mamm Genome 6:563–570. 6. Denny P, Lord CJ, Hill NJ, Goy JV, Levy ER, Podolin PL, Peterson LB, Wicker LS, Todd JA, Lyons PA (1997) Mapping
CYTOKINE, Vol 12, No. 5 (May, 2000: 477–482) of the IDDM locus Idd3 to a 0.35 cM interval containing the Interleukin-2 gene. Diabetes 46:695–700. 7. Smith KA (1992) Interleukin-2. Curr Opin Immunol 4: 271–276. 8. Hunig T, Schimpl A (1998) The IL-2 deficiency syndrome. A lethal disease caused by abnormal lymphocyte survival. In Durum SK, Muregge K (eds) Cytokine Knockouts. Humana Press Inc, Tolowa, pp 1–19. 9. Van Parijs L, Abbas AK (1998) Homeostasis and selftolerance in the immune system: Turning lymphocytes off. Science 280:243–248. 10. Ghosh S, Palmer SM, Rodrigues NR, Cordell HJ, Hearne CM, Cornall R, Prins J-B, McShane P, Lathrop GM, Peterson LB, Wicker LS, Todd JA (1993) Polygenic control of autoimmune diabetes in nonobese diabetic mice. Nature Genet 4:404–409. 11. Chesnut K, She J-X, Cheng I, Muralidharan K, Wakeland EK (1993) Characterizations of candidate genes for IDD susceptibility from the diabetes-prone NOD mouse strain. Mamm Genome 4:549–554. 12. Podolin PL, Denny P, Lord CJ, Hill NJ, Todd JA, Peterson LB, Wicker LS, Lyons PA (1997) Congenic mapping of the insulin dependent diabetes (Idd) gene, Idd10, localizes two genes mediating the Idd10 effect, and eliminates the candidate Fcgr1. J Immunol 159:1835–1843. 13. Podolin PL, Denny P, Armitage N, Lord CJ, Hill NJ, Levy ER, Peterson LB, Todd JA, Wicker LS, Lyons PA (1998) Localization of two insulin-dependent diabetes (Idd) genes to the Idd10 region on mouse chromosome 3. Mamm Genome 9:283–286. 14. Conradt HS, Geyer R, Hoppe J, Grotjahn L, Plessing A, Mohr H (1985) Structures of the major carbohydrates of natural human interleukin-2. Eur J Biochem 153:255–261. 15. Gillis S, Mochizuki DY, Conlon PJ, Hefeneider SH, Ramthun CA, Gillis AE, Frank MB, Henney CS, Watson JD (1982) Molecular characterization of interleukin 2. Immunol Rev 63: 167–209. 16. Robb RJ, Smith KA (1981) Heterogeneity of human T-cell growth factor(s) due to variable glycosylation. Mol Immunol 18:1087–1094. 17. Colomb E, Savino W, Wicker L, Peterson L, Dardenne M, Carnaud C (1996) Genetic control of giant perivascular space formation in the thymus of NOD mice. Diabetes 45:1535–1540. 18. Chen SL, Whiteley PJ, Freed DC, Rothbard JB, Peterson LB, Wicker LS (1994) Responses of NOD congenic mice to a glutamic acid decarboxylase-derived peptide. J Autoimmun 7: 635–641. 19. Kashima N, Nishi-Takaoka C, Fujita T, Taki S, Yamada G, Hamuro J, Taniguchi T (1985) Unique structure of murine interleukin-2 as deduced from cloned cDNAs. Nature 313:402–404. 20. Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of the bacteriophage T4. Nature 227: 680–685.