Abstract
Age-related macular degeneration (AMD) is the leading cause of blindness in the elderly population. Evidence has shown that the human immune system may play critical roles in this disease. Inducible costimulator (ICOS) promotes T-cell activation, differentiation, and T:B-cell interactions. The aim of the study was to understand the effect of ICOS on the development of AMD from genetic polymorphism perspective and serum level perspective. Two ICOS polymorphisms, rs10183087A/C and rs10932037C/T, were tested in 223 AMD cases and 262 healthy controls. The serum level of soluble ICOS (sICOS) was compared among subjects with different genotypes, as well as between AMD patients and controls. Data showed that prevalence of rs10183087CC genotype was significantly increased in AMD than in controls (p=0.001). Function analysis revealed that subjects carrying rs10183087CC genotype had higher serum levels of sICOS than those with AA or AC genotypes (p<0.05). When we compared serum levels of sICOS between cases and controls, results showed that AMD patients had significantly increased sICOS levels than healthy donors (p<0.05). Also, wet type cases were observed to have higher sICOS levels than cases with dry type (p<0.05). These data suggested ICOS polymorphism could affect the susceptibility to AMD by elevating protein expression, and serum levels of sICOS may be closed correlated with the development and progression of this disease.
Introduction
Age-related macular degeneration (AMD) is currently the leading cause of blindness in many countries and its prevalence will continue to increase in the coming decades unless new prevention strategies can be developed (Meyer et al., 2010). Currently, over 1.75 million U.S. citizens are affected with advanced AMD, and this number is likely to increase to nearly 3 million by the year 2020 due to the increased number of aging U.S. citizens (Meyer et al., 2010). The mechanisms leading to the development of both the early and late lesions of AMD remain largely unknown. Population studies have helped to determine risk factors for this disorder. Some of the nonmodifiable factors include age, genetic factors, light colored iris, and hyperopia (Coleman et al., 2008; Jager et al., 2008). Some of the modifiable risk factors include cigarette smoking, diet, high blood pressure, and possibly elevated serum cholesterol. Recent information supports the notion that immune mechanisms play an important and perhaps central role in AMD (Coleman et al., 2008; Jager et al., 2008).
Inducible costimulator (ICOS) is the third member of the CD28 family of costimulatory molecules, and is induced on the cell surface following T-cell activation (Swallow et al., 1999; Greenwald et al., 2005). It has been reported that the ICOS pathway promotes T-cell activation, differentiation and effector responses, and T-cell-dependent B-cell responses (McAdam et al., 2000; Coyle and Gutierrez-Ramos, 2004; Nurieva et al., 2008). ICOS mediated costimulation of T cells leads predominantly to the production of effector cytokines such as interleukin (IL)-4 and IL-10, and to a lesser extent, IL-2, interferon (IFN)-γ, and TNF-α (Lohning et al., 1998), thereby playing a more important role in Th2 responses than Th1 responses (Burmeister et al., 2008; Bauquet et al., 2009). Due to the importance of ICOS, aberrant ICOS expression has been shown to be associated with different diseases. ICOS surface expression has been found to be increased on T cells from patients with rheumatoid arthritis, and ligation of ICOS induces augmented IFN-γ production (Okamoto et al., 2003). Patients with systemic lupus erythrematosis were also found to have increased ICOS expression on their peripheral blood T cells (Hutloff et al., 2004).
ICOS may play critical roles in AMD. ICOS upregulates IL-10 and INF-γ, those have profound effects on AMD (Rudd et al., 2009). More importantly, ICOS can activate complement cascade, which is one of the main causes for the development of AMD (Jain et al., 2008; Sharpe, 2009). In the current study, we have investigated the role of ICOS on AMD from two different angles. First of all, we have tested ICOS rs10183087A/C and rs10932037C/T polymorphisms in AMD and analyzed their functions. Second, we have analyzed the serum levels of soluble ICOS (sICOS) in this disease.
Materials and Methods
Patients and controls
The study group included 223 AMD cases and 262 healthy controls recruited from the General Hospital of Guangzhou Military Command of PLA. The inclusion criteria for patients included those with age 50 years or older with the diagnosis of AMD. AMD was defined by geographic atrophy and/or choroidal neovascularization with drusen more than five in at least one eye. The exclusion criteria included the retinal diseases involving the photoreceptors and/or outer retinal layers other than AMD loss such as high myopia, retinal dystrophies, central serious retinopathy, vein occlusion, uveitis, or similar outer retinal diseases that have been present before the age of 50 and opacities of the ocular media, limitations of papillary dilation, or other problems sufficient to preclude adequate stereo fundus photography. The control group was recruited from people who underwent regular physical examinations at the same hospital. The study was approved by the Review Board of the General Hospital of Guangzhou Military Command of PLA. Each study participant provided a peripheral blood sample.
DNA extraction and genotyping
Genomic DNA was extracted from 5 mL frozen whole blood using the DNA Extraction Kit (Fastagen) according to the manufacturer's protocol. The ICOS rs10183087A/C and rs10932037C/T genotypes were determined using a polymerase chain reaction (PCR)–restriction fragment length polymorphism assay and DNA sequencing analysis. The PCR primers for the rs10183087A/C and rs10932037C/T polymorphisms were 5′-AGAGGGGET AL.,GCTTCTTGTAGGGA-3′ (forward) and 5′-CTCATCCCTACAGET AL.,GAAGAGGT-3′ (reverse), and 5′- CATTATCTATGTTTTCATGGTGCTATT-3′ (forward) and 5′- AGGCTATCTTGAAGGGCCAG-3′ (reverse), respectively. The PCRs were performed in a total volume of 25 μL containing 100 ng genomic DNA, 20 pM of each primer, 0.2 mM dNTPs, 20 mM Tris-HCl (pH 8.8), 10 mM KCl, 10 mM (NH4)2SO4, 2 mM MgSO4, 0.1%Triton X-100, and 1 unit of Taq polymerase (New England BioLabs). The PCR cycle conditions consisted of an initial denaturation step at 94°C for 5 min, followed by 35 cycles of 30 s at 94°C; 45 s at 61°C for rs10183087A/C and 63°C for rs10932037C/T; 45 s at 72°C; and a final elongation at 72°C for 8 min. The PCR products were digested for 3 h at 37°C with the appropriate restriction enzymes (New England BioLabs). The restriction enzymes for the rs10183087A/C and rs10932037C/T genotypes were Hpy188III and Hsp92II, respectively. The digested PCR products were resolved on 8% acrylamide gels and stained with ethidium bromide for visualization under UV light. As for the rs10183087, the fragment size of PCR was 532 bp. When conducting enzyme digestion, the fragment size indicating the A allele was 223 and 309, whereas the C allele was 90, 133, and 309. As for rs10932037, the fragment size of PCR was 445 bp. When conducting enzyme digestion, the fragment size indicating the C allele was 19, 76, 80, and 270, whereas the T allele was 19, 156, and 270. To confirm the genotyping results, more than 10% of PCR amplified DNA samples were examined by DNA sequencing. Results between PCR and DNA sequencing analysis were 100% concordant.
Measurement of serum sICOS levels
Serum levels of sICOS were measured in duplicate with a specific sandwich ELISA kit (Uscn Life Science, Inc.), according to the manufacturer's protocol.
Statistical analysis
Genotype and allele frequencies of ICOS polymorphisms were compared between AMD cases and controls using the chi-square test and odds ratios (OR), and 95% confidence intervals (CIs) were calculated to assess the relative risk conferred by a particular allele and genotype. Demographic and clinical data between groups were compared by the chi-square test and Student's t-test. Haplotypes of the SNPs were conducted and analyzed using the SHEsis software (Bio-X, Inc.). The serum level of sICOS was compared by the Student's t-test. Statistical significance was assumed at p<0.05. The SPSS statistical software package version 13.0 was used for all of the statistical analyses.
Results
Clinical characteristics of the study subjects
The clinical characteristics of all the subjects are shown in Table 1. There was no significant difference in age (p>0.05) and gender (p>0.05) between the AMD patients and the healthy adults. Compared to the control, the AMD group had higher levels of body mass index (BMI) (24.8±6.9), and higher prevalence of hypertension (18.8%) as well as diabetes (9.4%). However, none of these values showed significant differences. In the 223 AMD cases, 152 (68.2%) were wet type, whereas 71 (31.8%) were dry type.
Table 1.
General Characteristics of the Age-Related Macular Degeneration Patients and Control Group
| Characteristics | AMD (n=223) | Control (n=262) | p-Value |
|---|---|---|---|
| Age (years) | 66.7±8.2 | 68.3±7.6 | >0.05 |
| Gender (M/F) | 125/98 | 146/116 | >0.05 |
| BMI (kg/m2) | 24.8±6.9 | 23.3±4.9 | >0.05 |
| Hypertension, n (%) | 42 (18.8) | 38 (14.5) | >0.05 |
| Diabetes, n (%) | 21 (9.4) | 17 (6.5) | >0.05 |
| Wet AMD | 152 (68.2) | — | |
| Dry AMD | 71 (31.8) | — |
Data are presented as mean±standard deviation.
AMD, age-related macular degeneration; M/F, male/female; BMI, body mass index.
The ICOS polymorphisms in AMD cases and controls
The genotype and allele frequencies of the ICOS rs10183087A/C and rs10932037C/T polymorphisms in AMD cases and controls are summarized in Table 2. The genotype distributions of these two polymorphisms among the controls were in agreement with the Hardy–Weinberg equilibrium (p>0.05). The allele frequencies of ICOS rs10183087A/C polymorphism were 77.1% and 22.9% among the cases, and 86.3% and 13.7% among the controls (p=0.0002). Prevalence of the ICOS rs10183087CC genotype was significantly higher in patients than in controls (OR=3.81, 95% CI: 1.65–8.80, p=0.001). As for the rs10932037C/T polymorphism, the prevalence of T and G allele was 96.4% and 3.6% in patients and 97.1% and 2.9% in controls (p=0.522). We did not identify the ICOS rs10932037TT genotype in controls. The ICOS rs10932037CT genotype did not show any significant difference between the patients and the controls (OR=1.11, 95% CI: 0.52–2.35, p=0.788). In addition, we analyzed the haplotypes of the two polymorphisms. No significant linkage disequilibrium was observed among these SNPs. Haplotypes are shown (Table 2). The CC (rs10183087–rs10932037) haplotype had significantly higher numbers in AMD cases compared to healthy controls (OR=1.84, 95% CI: 1.31–2.59, P=0.0004). These data suggested that the ICOS rs10183087A/C polymorphism was associated with an increased susceptibility to AMD in the Chinese population.
Table 2.
ICOS Polymorphisms in Age-Related Macular Degeneration Patients and Controls
| Polymorphism | Cases n=223 (%) | Controls n=262 (%) | OR (95%CI) | p-Value |
|---|---|---|---|---|
| rs10183087 | ||||
| Genotype | ||||
| AA | 143 (64.1) | 198 (75.6) | 1.00 | |
| AC | 58 (26.0) | 56 (21.4) | 1.43 (0.94–2.20) | 0.096 |
| CC | 22 (9.9) | 8 (3.0) | 3.81 (1.65–8.80) | 0.001a |
| Allele | ||||
| A | 344 (77.1) | 452 (86.3) | 1.00 | |
| C | 102 (22.9) | 72 (13.7) | 1.86 (1.34–2.60) | 0.0002a |
| rs10932037 | ||||
| Genotype | ||||
| CC | 208 (93.3) | 247 (94.3) | 1.00 | |
| CT | 14 (6.3) | 15 (5.7) | 1.11 (0.52–2.35) | 0.788 |
| TT | 1 (0.4) | 0 (0.0) | — | — |
| Allele | ||||
| C | 430 (96.4) | 509 (97.1) | 1.00 | |
| T | 16 (3.6) | 15 (2.9) | 1.26 (0.62–2.58) | 0.522 |
| Haplotypes | ||||
| rs10183087– rs10932037 | ||||
| AC | 335 (75.1) | 441 (84.2) | 1.00 | |
| CC | 95 (21.3) | 68 (13.0) | 1.84 (1.31–2.59) | 0.0004a |
| AT | 9 (2.0) | 11 (2.1) | 1.08 (0.44–2.63) | 0.870 |
| CT | 7 (1.6) | 4 (0.7) | 2.30 (0.67–7.94) | 0.174 |
p-Value is statistically significant (<0.05).
ICOS, inducible costimulator; OR, odds ratio; CI, confidence interval.
Comparison of ICOS polymorphisms on protein expression
To learn the mechanism of the association between ICOS polymorphisms and AMD, we first compared the effect of rs10183087A/C polymorphism on the serum level of sICOS in healthy controls (Fig. 1A). To avoid the potential interference from rs10932037C/T polymorphism, healthy donors carrying this genetic variant were excluded. Data showed that the serum level of subjects with wild-type rs10183087AA genotype was 466±98 pg/mL (mean±standard deviation), whereas subjects with AC and CC genotypes were 496±107 pg/mL and 704±108 pg/mL, respectively. Subjects with rs10183087CC genotype had a significantly higher level of sICOS than those with wild-type AA and AC genotypes (p<0.05) (Fig. 1A). We also compared the effect of rs10932037C/T polymorphism on the serum level of sICOS in healthy controls (Fig. 1A). Similarly, subjects having rs10183087A/C polymorphism were excluded to avoid interference. Results revealed that serum levels of subjects with rs10932037CC and rs10932037CT genotypes were 449±91 pg/mL and 478±111 pg/mL, respectively. No significant differences were observed between CC and CT genotypes (Fig. 1A). By using similar methods, we further compared the effect of these two polymorphisms on the serum level of sICOS in AMD patients (Fig. 1B). Patients with rs10183087CC genotype demonstrated elevated levels of sICOS than patients with AA and AC genotypes (p<0.05), whereas rs10932037C/T polymorphism did not present any effect on the sICOS levels in AMD patients (Fig. 1B). These results indicated that the rs10183087A/C polymorphism could increase the sICOS serum level in both healthy subjects and AMD cases.
FIG. 1.
Serum level of soluble inducible costimulator (sICOS) in the study population with different polymorphisms. (A) Shown in healthy controls. (B) Shown in age-related macular degeneration (AMD) patients. Data are presented as mean±standard deviation (SD).
Serum level of sICOS in AMD cases and controls
To understand the effect of ICOS on AMD, we compared the serum level of sICOS in AMD cases and controls (Fig. 2A). Subjects with ICOS polymorphisms were excluded to avoid interference. As shown in Figure 2A, an increased sICOS expression was observed in AMD patients (597±99 pg/mL) than in healthy controls (466±98 pg/mL) (p<0.05). To learn whether the ICOS level is correlated with some clinical parameters, we analyzed serum ICOS in patients with different hypertension status (Fig. 2B), diabetes status (Fig. 2C), or vessel affected status (Fig. 2D). Data showed that the serum ICOS level was not associated with hypertension or diabetes in AMD (p>0.05). However, AMD patients who had wet type presented significantly higher sICOS levels than those with dry type (700±150 pg/mL versus 494±56 pg/mL, p<0.05) (Fig. 2D). These results indicated that serum sICOS may play roles in the progression of AMD.
FIG. 2.
Comparison of serum sICOS between controls and AMD patients (A). Comparison of serum sICOS levels between AMD cases with hypertension and AMD cases without hypertension (B); between AMD cases with diabetes and AMD cases without diabetes (C); between cases with wet type and cases with dry type (D). Data are presented as mean±SD.
Discussion
Evidence is accumulating that the human immune system is closely correlated with the development of AMD (Coleman et al., 2008; Jager et al., 2008). However, factors responsible for the pathogenesis of AMD remain unclear. The current study investigated the effect of ICOS on AMD from various angles. We first identified that ICOS rs10183087A/C polymorphism could affect the susceptibility to AMD by increasing protein expression. Further, we found that the serum sICOS level was elevated in AMD patients and associated with wet type. This study demonstrated that ICOS may play critical roles in the development and progression of AMD.
ICOS rs10183087A/C and rs10932037C/T SNPs were located at the 3′ untranslated region (3′UTR) (Haimila et al., 2002), by which the gene expression and mRNA stability may be influenced (Wu et al., 2009). Studies have presented that polymorphisms of ICOS may be associated with various diseases. It has been reported that recipients of hematopoietic stem cell transplantation who received a graft from a donor with ICOS rs10183087CC genotype had worse disease-free survival (Wu et al., 2009). In addition, Castelli et al. (2007) investigated eight SNPs at the ICOS 3′UTR and mentioned that the homozygote of haplotype A (602A–930G–1459A–1564T–1624C–1862A–2007G–2373G) decreased ICOS expression and might play a protective role in multiple sclerosis. Our data have shown that the ICOS rs10183087CC genotype can significantly increase the serum level of sICOS in healthy controls (Fig. 1A), suggesting a possible function of this polymorphism in gene expression. Similarly, AMD patients with the ICOS rs10183087CC genotype also showed higher sICOS levels, indicating that the effect of ICOS rs10183087A/C polymorphism on protein expression is ubiquitous, or at least is not restricted in healthy people. As for the rs10932037C/T SNP, previous research has reported that recipients of hematopoietic stem cell transplantation with the ICOS rs10932037TT genotype had an overall worse survival. Our study did not observe any changes in sICOS levels with regard to the polymorphism (Fig. 1), suggesting that it may not affect the gene expression at the protein level.
Increased ICOS expression has been associated with disease in several human studies. ICOS surface expression has been found to be increased on T cells from patients with rheumatoid arthritis, and ligation of ICOS induces augmented INF-γ production (Okamoto et al., 2003). Patients with systemic lupus erythrematosis were also found to have increased ICOS expression on their peripheral blood T cells (Hutloff et al., 2004). The increase in ICOS expression correlated with increased IL-4, IL-5, and IL-13 production from ex vivo stimulated peripheral blood mononuclear cells. In addition, studies have suggested that the level of ICOS expression may contribute to the proliferation and migration of Tfh (Shilling et al., 2006). In the current study, we observed elevated sICOS levels in AMD patients, especially in patients with wet type. The mechanism remains unclear yet. Since ICOS regulates multiple cytokine productions, whereas cytokines may play critical roles in the development of AMD, we hypothesize that ICOS may affect AMD indirectly through cytokines. In addition, we identified that AMD with wet type had higher sICOS levels than those with dry type (Fig. 2D), suggesting that sICOS may be closely involved in the development of the disease. Further studies on this field are necessary.
In summary, our study identified that ICOS polymorphism could affect the susceptibility to AMD by increasing ICOS expression, and found that the serum sICOS level was elevated in AMD patients, especially in patients with wet type. This research sheds light on understanding the pathogenesis of AMD and provides therapeutic strategies in this disease.
Acknowledgment
This study was supported by Grant 2011KF03 from the State Key Laboratory of Ophthalmology, Zhongshan Ophthalmic Center, Sun Yat-sen University, Guangzhou 510060, China.
Disclosure Statement
No competing financial interests exist.
References
- Bauquet A.T. Jin H. Paterson A.M. Mitsdoerffer M. Ho I.C. Sharpe A.H. Kuchroo V.K. The costimulatory molecule ICOS regulates the expression of c-Maf and IL-21 in the development of follicular T helper cells and TH-17 cells. Nat Immunol. 2009;10:167–175. doi: 10.1038/ni.1690. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Burmeister Y. Lischke T. Dahler A.C. Mages H.W. Lam K.P. Coyle A.J. Kroczek R.A. Hutloff A. ICOS controls the pool size of effector-memory and regulatory T cells. J Immunol. 2008;180:774–782. doi: 10.4049/jimmunol.180.2.774. [DOI] [PubMed] [Google Scholar]
- Castelli L. Comi C. Chiocchetti A. Nicola S. Mesturini R. Giordano M. D'Alfonso S. Cerutti E. ICOS gene haplotypes correlate with IL10 secretion and multiple sclerosis evolution. J Neuroimmunol. 2007;186:193–198. doi: 10.1016/j.jneuroim.2007.03.022. [DOI] [PubMed] [Google Scholar]
- Coleman H.R. Chan C.C. Ferris F.L., 3rd Chew E.Y. Age-related macular degeneration. Lancet. 2008;372:1835–1845. doi: 10.1016/S0140-6736(08)61759-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Coyle A.J. Gutierrez-Ramos J.C. The role of ICOS and other costimulatory molecules in allergy and asthma. Springer Semin Immunopathol. 2004;25:349–359. doi: 10.1007/s00281-003-0154-y. [DOI] [PubMed] [Google Scholar]
- Greenwald R.J. Freeman G.J. Sharpe A.H. The B7 family revisited. Annu Rev Immunol. 2005;23:515–548. doi: 10.1146/annurev.immunol.23.021704.115611. [DOI] [PubMed] [Google Scholar]
- Haimila K.E. Partanen J.A. Holopainen P.M. Genetic polymorphism of the human ICOS gene. Immunogenetics. 2002;53:1028–1032. doi: 10.1007/s00251-002-0431-2. [DOI] [PubMed] [Google Scholar]
- Hutloff A. Büchner K. Reiter K. Baelde H.J. Odendahl M. Jacobi A. Dörner T. Kroczek R.A. Involvement of inducible costimulator in the exaggerated memory B cell and plasma cell generation in systemic lupus erythematosus. Arthritis Rheum. 2004;50:3211–3220. doi: 10.1002/art.20519. [DOI] [PubMed] [Google Scholar]
- Jain V.V. Perkins D.L. Finn P.W. Costimulation and allergic responses: immune and bioinformatic analyses. Pharmacol Ther. 2008;117:385–392. doi: 10.1016/j.pharmthera.2007.12.002. [DOI] [PubMed] [Google Scholar]
- Jager R.D. Mieler W.F. Miller J.W. Age-related macular degeneration. N Engl J Med. 2008;358:2606–2617. doi: 10.1056/NEJMra0801537. [DOI] [PubMed] [Google Scholar]
- Lohning M. Hutloff A. Kallinich T. Mages H.W. Bonhagen K. Radbruch A. Hamelmann E. Kroczek R.A. Expression of ICOS in vivo defines CD4+ effector T cells with high inflammatory potential and a strong bias for secretion of interleukin 10. J Exp Med. 1998;197:181–193. doi: 10.1084/jem.20020632. [DOI] [PMC free article] [PubMed] [Google Scholar]
- McAdam A.J. Chang T.T. Lumelsky A.E. Greenfield E.A. Boussiotis V.A. Duke-Cohan J.S. Mouse inducible costimulatory molecule (ICOS) expression is enhanced by CD28 costimulation and regulates differentiation of CD4+ T cells. J Immunol. 2000;165:5035–5040. doi: 10.4049/jimmunol.165.9.5035. [DOI] [PubMed] [Google Scholar]
- Meyer K.J. Davis L.K. Schindler E.I. Beck J.S. Rudd D.S. Grundstad A.J. Genome-wide analysis of copy number variants in age-related macular degeneration. Hum Genet. 2010;129:91–100. doi: 10.1007/s00439-010-0904-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nurieva R.I. Chung Y. Hwang D. Yang X.O. Kang H.S. Ma L. Wang Y.H. Generation of T follicular helper cells is mediated by interleukin-21 but independent of T helper 1, 2, or 17 cell lineages. Immunity. 2008;29:138–149. doi: 10.1016/j.immuni.2008.05.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Okamoto T. Saito S. Yamanaka H. Tomatsu T. Kamatani N. Ogiuchi H. Uchiyama T. Yagi J. Expression and function of the co-stimulator H4/ICOS on activated T cells of patients with rheumatoid arthritis. J Rheumatol. 2003;30:1157–1163. [PubMed] [Google Scholar]
- Rudd C.E. Taylor A. Schneider H. CD28 and CTLA-4 coreceptor expression and signal transduction. Immunol Rev. 2009;229:12–26. doi: 10.1111/j.1600-065X.2009.00770.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sharpe A.H. Mechanisms of costimulation. Immunol Rev. 2009;229:5–11. doi: 10.1111/j.1600-065X.2009.00784.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shilling R.A. Bandukwala H.S. Sperling A.I. Regulation of T:B cell interactions by the inducible costimulator molecule: does ICOS “induce” disease? Clin Immunol. 2006;121:13–18. doi: 10.1016/j.clim.2006.04.574. [DOI] [PubMed] [Google Scholar]
- Swallow M.M. Wallin J.J. Sha W.C. B7h, a novel costimulatory homolog of B7.1 and B7.2, is induced by TNFalpha. Immunity. 1999;11:423–432. doi: 10.1016/s1074-7613(00)80117-x. [DOI] [PubMed] [Google Scholar]
- Wu J. Tang J.L. Wu S.J. Lio H.Y. Yang Y.C. Functional polymorphism of CTLA-4 and ICOS genes in allogeneic hematopoietic stem cell transplantation. Clin Chim Acta. 2009;403:229–233. doi: 10.1016/j.cca.2009.03.037. [DOI] [PubMed] [Google Scholar]