SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES

Objective

Complement component C4 copy number variations are associated with various inflammatory diseases. This study was undertaken to investigate whether copy number variations of C4 are also involved in the pathogenesis of Behçet's disease (BD).

Methods

Gene expression was examined by enzyme-linked immunosorbent assay (ELISA) or real-time polymerase chain reaction (PCR). Copy number variations of C4 isotypes (C4A and C4B) were detected by real-time PCR in 905 patients with BD, 205 patients with ankylosing spondylitis (AS) and acute anterior uveitis, and 1,238 controls. The activation of CD4+ T cells was analyzed by flow cytometry, and cytokine production was detected by ELISA.

Results

Protein expression of total C4 in serum was significantly increased in patients with active BD compared with those with inactive BD or controls (Bonferroni corrected P [Pcorr] = 1.64 × 10−4 and Pcorr = 0.037, respectively), but not in patients with AS and acute anterior uveitis. Copy number variation analysis identified a significantly increased frequency of more than 2 copies of C4A in BD patients (P = 1.65 × 10−7, odds ratio [OR] 2.84). HLA–B51, which is located on the same chromosome as C4, showed a strong association with BD in the Han Chinese population (P = 8.90 × 10−65, OR 5.05), but logistic regression showed that C4A copy number variation was an independent risk factor for BD. A significantly increased expression of C4A was observed in the high copy number groups (>2 copies or 2 copies) versus the low copy number group (Pcorr = 0.019 and Pcorr = 0.044, respectively). Increased production of interleukin-6 (IL-6) was also observed in the high C4A copy number group (Pcorr = 0.037). No effect of C4 copy number variation on the expression of T cell activation markers was detected.

Conclusion

Our findings indicate that a high copy number of C4A confers risk for BD by modulating the expression of C4A and enhancing IL-6 production.

Uveitis is one of the main causes of blindness worldwide ([1]). Behçet's disease (BD) and acute anterior uveitis associated with ankylosing spondylitis (AS) are two commonly seen uveitis entities in Asia ([2-4]). However, the pathogenesis and clinical features of these diseases are quite different. BD is a chronic systemic inflammatory disorder characterized by nongranulomatous uveitis, recurrent oral ulcers, genital ulcers, and skin lesions ([5]). AS is a chronic autoimmune arthritis that primarily affects the spine and is characterized by inflammatory back pain, asymmetrical peripheral oligoarthritis, and enthesitis ([6]). Acute anterior uveitis is considered to be the most frequent extraarticular manifestation of AS ([7]). Although the exact etiology of uveitis remains unclear, host genetic factors as well as immune imbalance are potential cofactors implicated in the initiation and development of the disease.

The complement system is seen as a major element of innate immunity and is regarded as the first line of defense against intrinsic and extrinsic antigens. It has been shown to be involved in the development of several ocular inflammatory diseases ([8-10]). Furthermore, a variety of genetic variants of the complement system are associated with ocular inflammatory diseases, such as age-related macular degeneration ([11-13]) and anterior uveitis ([14, 15]). Wakefield et al ([14]) were the first to address a possible role for complement C4 allotypes in uveitis and found a significantly increased frequency of the C4B2 allotype in patients with anterior uveitis. C4, which is located on chromosome 6q21.3, encodes 2 functionally distinct C4A (MIM 120810) and C4B (MIM 120820) isoforms. Previous studies indicated that a low copy number of C4A or C4B confers risk for multiple autoimmune disorders, including systemic lupus erythematosus (SLE) ([16-18]), rheumatoid arthritis (RA) ([19]), and Graves' disease ([20]). A positive correlation between gene copy number variations of C4 and the quantity of C4 protein has been demonstrated ([21]). Additionally, specific C4 deficiency has been associated with multiple autoimmune disorders, including SLE ([22]), RA ([23]), and juvenile idiopathic arthritis ([24]).

The aforementioned data suggest that copy number variations of C4 are associated with multiple immune diseases and may contribute to C4 deficiency in patients with SLE or RA. Whether C4 levels are aberrant in uveitis patients, including patients with BD and patients with AS-associated acute anterior uveitis, and whether copy number variations of C4 are responsible for an altered level in uveitis has not yet been reported. We therefore analyzed the level of C4 and C4 copy number variations in two frequently encountered clinical uveitis entities in China and investigated associations between the variants and a number of immune response parameters. Our findings provide further support for the notion that C4 plays an important role in the pathogenesis of intraocular inflammation.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES

Recruitment of patients and normal controls

A total of 905 patients with BD, 205 patients with AS and acute anterior uveitis, and 1,238 healthy individuals were recruited from The First Affiliated Hospital of Chongqing Medical University (Chongqing, China) or the Zhongshan Ophthalmic Center, Sun Yat-sen University (Guangzhou, China) (details are available from the corresponding author upon request). The diagnoses of BD and AS were based strictly on the criteria of the International Study Group for BD ([25]) and the 1984 modified New York Criteria for AS ([26]), respectively. If the diagnosis was in doubt, the patients were excluded from the study. The controls used in this study were unrelated healthy individuals without any autoimmune disorder. They were matched with the patients for age and ethnicity. All study participants were Han Chinese. Written informed consent was obtained from all subjects. The study was approved by the Ethics Committee of our hospitals (permit number 2009-201008) and was conducted in accordance with the Declaration of Helsinki.

Genomic DNA extraction and analysis of complement C4 gene copy number variations

Genomic DNA was extracted from peripheral blood using a Qiagen QIAamp DNA Mini blood kit. C4 gene copy number variations were determined by TaqMan real-time polymerase chain reaction (PCR) and performed in 96-well optical plates on an Applied Biosystems 7500 real-time PCR system following the manufacturer's protocols. TaqMan assays labeled with FAM were used to detect C4A (Hs07226349-cn) and C4B (Hs07226350-cn), respectively (Applied Biosystems). TaqMan RNase P assay labeled with VIC was used as an internal copy number reference (Applied Biosystems).

HLA–B51 genotyping

HLA–B51 allelic typing was performed using the PCR sequence-specific primer method as previously described ([27, 28]).

Cell isolation and culture

Peripheral blood mononuclear cells (PBMCs) were prepared from venous blood by Ficoll-Hypaque density-gradient centrifugation. Isolated PBMCs were resuspended at a concentration of 1 × 106 cells/ml in RPMI 1640 complete medium supplemented with 10% fetal calf serum, 2 mML-glutamine, and 100 units/ml penicillin/streptomycin (Invitrogen). PBMCs were treated with lipopolysaccharide (100 ng/ml; Sigma) for 24 hours to stimulate tumor necrosis factor α (TNFα), interleukin-1β (IL-1β), IL-6, IL-8, and monocyte chemotactic protein 1 (MCP-1) secretion in the culture supernatants. To stimulate IL-17, interferon-γ (IFNγ), and IL-10 production, PBMCs were treated with a combination of anti-CD3 (OKT3; 0.5 μg/ml) and anti-CD28 antibodies (15E8; 0.1 μg/ml) (Miltenyi Biotec) for 72 hours.

Real-time quantitative PCR analysis

Total RNA was isolated from PBMCs using a QIAamp RNA Blood Mini kit (Qiagen) with DNase I treatment according to the manufacturer's instructions, and reverse transcribed into complementary DNA (cDNA) according to the Superscript protocol (SuperScript III first-strand synthesis system; Invitrogen). Real-time quantitative PCR was performed using an Applied Biosystems 7500 real-time PCR system. The expression of C4A was examined using the following primers: forward 5′-GGACCCCTGTCCAGTGTTAG-3′ and reverse 5′-GATGAAGGGCGATGGTCACA-3′. β-actin was chosen as the internal reference gene to normalize the expression of target genes as previously described ([29, 30]). The inline image method was applied to quantify the relative levels of target genes using the following formulas: inline image and inline image . A pooled sample of cDNA from all normal controls was used as a calibrator.

Flow cytometric analysis

The expression of CD69 and CD25 activation markers was analyzed using flow cytometry. PBMCs were incubated for 30 minutes at 4°C with fluorescein isothiocyanate–conjugated anti-human CD4, allophycocyanin- conjugated anti-human CD44, phycoerythrin (PE)–conjugated anti-human CD25, PE–Cy7–conjugated anti-human CD69, or appropriate isotypes (eBioscience). Flow cytometric analysis was performed on a FACSAria cytometer (BD Biosciences). Results were analyzed using FlowJo software (Tree Star).

Measurement of C4 and cytokine levels by enzyme-linked immunosorbent assay (ELISA).

Serum from patients and controls was stored at −80°C. The C4 level was measured using a human ELISA kit (Cusabio Biotech). Supernatants of stimulated PBMCs were collected and stored at −80°C until cytokine levels were measured. The concentrations of TNFα, IL-1β, IL-6, IL-8, MCP-1, IL-17, IFNγ, and IL-10 were determined using DuoSet ELISA development kits according to the recommendations of the manufacturer (R&D Systems).

Statistical analysis

Real-time PCR data were analyzed using Applied Biosystems 7500 system software, version 2.0.6. Relative gene copy numbers were examined by the comparative Ct method using CopyCaller version 2.0 (Applied Biosystems). The number of C4A and C4B genes equals the number of total C4 copies in an individual ([17]). The differences in total C4, C4A, and C4B were compared between patients and controls using copy number variations of <2 or <4 as a reference (odds ratio [OR] 1) by chi-square test using SPSS version 17.0 software. Logistic regression analysis was also performed to evaluate the association between C4 copy number variations and phenotype controlling for the effect of HLA–B51. The expression of C4, C4A, and various cytokines was analyzed by t-test for independent samples or nonparametric test for 2 independent samples, using SPSS version 17.0. To account for multiple testing, Bonferroni correction was applied.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES

Clinical findings in patients with BD and patients with AS and acute anterior uveitis

The distribution of demographic characteristics and clinical characteristics of the enrolled patients with BD and patients with AS and acute anterior uveitis were assessed at the time of diagnosis (details are available from the corresponding author upon request). The mean ± SD age of the normal controls was 40.6 ± 10.7 years.

Increased serum C4 protein levels in patients with active BD

Previous studies have shown an association between altered C4 levels and various autoimmune or autoinflammatory diseases ([31, 32]). AS-associated acute anterior uveitis is considered an autoimmune disease, whereas BD is seen as an autoinflammatory disease ([33]). We investigated whether the protein expression of total C4 in serum obtained from patients with AS-associated acute anterior uveitis or patients with BD was different from that seen in normal controls. Serum samples were obtained from 39 patients with active BD, 32 patients with inactive BD, 13 patients with active AS-associated acute anterior uveitis, 17 patients with inactive AS-associated acute anterior uveitis, and 26 healthy controls, and protein expression of total C4 was assessed. The expression of total C4 was significantly increased in patients with active BD as compared with normal controls or patients with inactive BD (Bonferroni corrected P [Pcorr] = 0.037 and Pcorr = 1.64 × 10−4, respectively) (Figure 1). However, no significant difference was found between patients with inactive BD, patients with active AS-associated acute anterior uveitis, or patients with inactive AS-associated acute anterior uveitis and healthy controls (P > 0.05 for each comparison) (Figure 1).

image

Figure 1. C4 levels in serum as measured by enzyme-linked immunosorbent assay. Serum samples were obtained from 26 normal controls (NC), 32 patients with inactive Behçet's disease (iBD), 39 patients with active BD (aBD), 17 patients with inactive ankylosing spondylitis and acute anterior uveitis (iAS), and 13 patients with active AS and acute anterior uveitis (aAS), and the expression of total C4 was measured. Symbols represent individual subjects; horizontal lines show the mean. Corrected P values (Pc) were determined by nonparametric test for 2 independent samples with Bonferroni correction for 6 comparisons, using SPSS software.

Download figure to PowerPoint

Increased frequency of high C4A copy numbers in patients with BD

Accumulating evidence is available to show that the expression of genes can be influenced by their copy numbers ([21]). We therefore performed copy number variation analysis of C4 and C4 isotypes (C4A and C4B) in patients with uveitis and controls. We found a significantly increased frequency of having 2 copies or more than 2 copies of C4A in BD patients versus controls (P = 3.27 × 10−5, OR 2.23 [95% confidence interval 1.52–3.28] and P = 1.65 × 10−7, OR 2.84 [95% confidence interval 1.90–4.24], respectively) (Table 1). The frequency of having more than 2 copies of C4B was significantly decreased in BD patients as compared with controls (P = 5.14 × 10−3, OR 0.65 [95% confidence interval 0.48–0.88]). The significance of this latter association was lost after Bonferroni correction (n = 12; Pcorr = 0.062) (Table 1).

Table 1. Comparison of gene copy numbers of C4A, C4B, and total C4 among patients with BD, patients with AS, and unrelated controls*
 Copy number, no. of subjects<2 copies, OR (95% CI)a2 copiesa>2 copiesa
012345≥56≥7OR (95% CI)PbOR (95% CI)Pb
  1. BD = Behçet's disease; AS = ankylosing spondylitis.

  2. a

    The odds ratios (ORs), 95% confidence intervals (95% CIs), and P values shown for total C4 are for <4 copies, 4 copies, and >4 copies.

  3. b

    P values less than 0.05 (or P values with Bonferroni correction for 12 comparisons of less than 4.17 × 10−3) were considered significant.

C4A              
Controls (n = 1,234)51107842743427     
BD (n = 905)1365622176623Reference2.23 (1.52–3.28)3.27 × 10−52.84 (1.90–4.24)1.65 × 10−7
AS (n = 205)01612943143Reference1.18 (0.68–2.06)0.551.29 (0.71–2.32)0.40
C4B              
Controls (n = 1,238)152157981524117     
BD (n = 905)1417161091172Reference0.95 (0.76–1.19)0.650.65 (0.48–0.88)5.14 × 10−3
AS (n = 204)4371233631Reference0.87 (0.59–1.27)0.461.07 (0.67–1.72)0.78
Total C4              
Controls (n = 1,232)61587162367541     
BD (n = 905)61015101916532Reference1.09 (0.83–1.43)0.521.25 (0.94–1.67)0.13
AS (n = 204)42910838187Reference0.75 (0.49–1.15)0.180.89 (0.56–1.41)0.62

Consistent with the findings of our previous study ([34]), HLA–B51 was strongly associated with BD in Han Chinese (P = 8.90 × 10−65, OR 5.05 [95% confidence interval 4.16–6.12]). Since HLA–B51 and C4 are located on the same chromosome, we performed logistic regression to determine the independence of C4 copy number variations from the HLA–B51 association. The results showed that C4A copy number variation was an independent risk factor for BD in our population (for 2 copies of C4A, P = 2.36 × 10−3, OR 1.90 and for >2 copies of C4A, P = 9.17 × 10−4, OR 2.08) (Table 2). No association was found between copy number variations of C4A or C4B and AS-associated acute anterior uveitis. There was no association between total C4 copy number and BD or AS-associated acute anterior uveitis (Table 1).

Table 2. Logistic analysis of C4A in patients with BD and unrelated controls, controlling for the effect of HLA–B51*
 Copy number, no. of subjects<2 copies, OR (95% CI)†2 copies>2 copies
<22>2OR (95% CI)POR (95% CI)P
  1. BD = Behçet's disease; OR = odds ratio; 95% CI = 95% confidence interval.

Controls (n = 1,234)        
HLA–B51 positive1316485     
HLA–B51 negative102620250     
BD (n = 905)   Reference1.90 (1.26–2.87)2.36 × 10−32.08 (1.35–3.20)9.17 × 10−4
HLA–B51 positive19295187     
HLA–B51 negative18267119     

Association of clinical findings in patients with BD with C4 copy number variations

Since BD has a variety of ocular or extraocular manifestations, including hypopyon, skin lesions, genital ulcers, arthritis, and positive pathergy test results, we performed a stratified analysis to examine the association between C4 copy number variations and BD with or without these clinical parameters. There was no association between total C4, C4A, or C4B copy number variations and the aforementioned clinical parameters (P > 0.05) (details are available from the corresponding author upon request).

Up-regulated expression of C4A in healthy individuals carrying high copy numbers of C4.

The results described above indicated that copy number variations of C4A were associated with BD. We subsequently performed a functional analysis to assess the effect of C4A copy number variations on C4A expression (Figure 2). Significantly increased expression of C4A was observed in the high C4A copy number group (>2 copies of C4A) as compared with the low copy number group (<2 copies of C4A) (Pcorr = 0.019) (Figure 2).

image

Figure 2. Influence of copy number variations in C4A on the expression of C4A. The expression of C4A in peripheral blood mononuclear cells from normal controls carrying different copy number of C4A is shown. Symbols represent individual subjects (n = 6 for <2 copies, n = 18 for 2 copies, and n = 9 for >2 copies); horizontal lines show the mean. Corrected P values (Pc) were determined by nonparametric test for 2 independent samples with Bonferroni correction for 3 comparisons, using SPSS software.

Download figure to PowerPoint

Influence of copy number variations on the expression of CD4+ T cell activation markers

Previous studies showed that C4-knockout mice accumulate activated T cells ([35]). To study the role of C4 copy number variations on T cell activation, we examined the early and late activation markers of CD4+ T cells in different C4A copy number variation groups (Figure 3). The results showed that C4 copy number variation did not affect the expression of either CD69 or CD25 activation markers (Figure 3).

image

Figure 3. Influence of copy number variations in C4A on the activation of CD4+ T cells. The expression of the CD69 and CD25 activation markers in peripheral blood mononuclear cells from normal controls carrying different copy numbers of C4A is shown. Symbols represent individual subjects (n = 3 for <2 copies, n = 24 for 2 copies, and n = 8 for >2 copies); horizontal lines show the mean. Significance was examined by t-test for independent samples with Bonferroni correction for 6 comparisons, using SPSS software.

Download figure to PowerPoint

Influence of copy number variations of C4A on cytokine production

A number of studies conducted in recent years have indicated that the cytokine network represented by TNFα, IL-1β, IL-6, IL-8, MCP-1, IL-17, IL-10, and IFNγ plays an important role in the pathogenesis of BD ([36]). We therefore designed further experiments to investigate whether the different copy number variations of C4A affected the production of these cytokines (Figure 4). The results revealed that the group with a high C4A copy number (>2 copies) had an increased production of IL-6 (Pcorr = 0.037) as compared to the group with a low C4A copy number (<2 copies) (Figure 4E). No significant association was observed between IFNγ, IL-17, IL-10, IL-1β, IL-8, MCP-1, or TNFα production by stimulated PBMCs and C4A copy number (Figures 4A–D and F–H).

image

Figure 4. Influence of copy number variations in C4A on cytokine production. The levels of A, interferon-γ (IFNγ), B, interleukin-17 (IL-17), C, IL-10, D, IL-1β, E, IL-6, F, IL-8, G, monocyte chemotactic protein 1 (MCP-1), and H, tumor necrosis factor α (TNFα) in peripheral blood mononuclear cells from normal controls carrying different copy numbers of C4A are shown. Symbols represent individual subjects (n = 5 for <2 copies, n = 45 for 2 copies, and n = 22 for >2 copies); horizontal lines show the mean. Significance was examined by nonparametric test for 2 independent samples for IL-8 and TNFα and by t-test for independent samples for IFNγ, IL-17, IL-10, IL-1β, IL-6, and MCP-1, with Bonferroni correction for 24 comparisons, using SPSS software.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES

In the present study, we showed that the protein level of C4 in serum was increased in patients with active BD as compared to healthy controls, whereas no difference was observed between controls and the patients with AS and acute anterior uveitis. We furthermore found that having 2 copies or more than 2 copies of the C4A gene significantly increased the risk of BD. Functional studies showed significantly increased expression of C4A messenger RNA by PBMCs in the high copy number group. Increased production of IL-6 was found in the high C4A copy number group. We did not detect an effect of C4 copy number variations on the activation of CD4+ T cells.

Complement C4 plays several effector roles in the immune system, including involvement in the activation of the classical and the lectin complement pathways, which leads to cytolysis or neutralization of invading microbes, clearance of immune complexes, removal of apoptotic cells, and regulation of the activation of T and B lymphocytes ([35, 37]). Consistent with the findings of a previous study showing an increased level of C4 in patients with active BD ([32]), the present study showed increased expression of C4 in patients with active BD, but not in patients with AS-associated acute anterior uveitis.

C4 is cleaved into C4A and C4B by a subunit of the first complement component, C1s. Activated C4A will usually form an amide bond with amino groups on antigens, and it thereby plays a role in immune complex clearance, whereas activated C4B is strongly reactive with hydroxyl groups on glycerols or glycosylated antigens and is more relevant in the defense against microbes ([18, 38]). Furthermore, C4A may mediate an inflammatory response by stimulating smooth muscle contraction, histamine release from mast cells, vasodilation, and increased vascular permeability ([39]), leading to disease manifestations. The involvement of C4A in BD demonstrated in the present study, combined with data on vascular permeability changes ([40]), may provide another view to strengthen the hypothesis that the pathogenesis of BD includes a systemic vascular inflammation caused by the generation of immune complexes ([41]). Interestingly, we also found that the frequency of high copy numbers of C4A was increased in BD and that a higher copy number of C4A was associated with a higher expression of C4A. Our findings are consistent with those of an earlier study that showed a positive relationship between gene copy number variations of C4 and C4 protein quantity ([21]). These findings suggest that C4A copy number variations may contribute to the increased C4 level in BD.

The role of C4 in an autoimmune disease such as SLE seems to differ from the role of C4 in BD indicated by our findings. Hauptmann et al ([42]) were the first to report a link between C4 deficiency and SLE. Animal studies with mice also confirmed that C4 deficiency can lead to SLE ([43]). Immune complexes are considered to play an important role in the pathogenesis of SLE, and C4 deficiency may impair the clearance of these complexes, leading to tissue deposition. Low copy numbers of C4A or C4B have been shown to be more frequent in patients with SLE ([16-18]), RA ([19]), and Graves' disease ([20]). How C4 copy number affects RA or Graves' disease is unknown. Taken together, these data suggest that the role of C4 copy number variations may vary among immune-mediated diseases in humans.

We performed a number of experiments to assess a possible role for C4 copy number in cytokine expression and T cell activation. We performed these studies using samples from healthy controls, since the patients included in the present study represented a heterogeneous group of individuals with marked differences in the degree of inflammation and immunosuppressive therapy received. Our results showed that high copy numbers of C4A enhanced the production of IL-6. Despite the lack of a direct association between C4A and IL-6 expression, previous studies have found simultaneously elevated levels of C4 and IL-6 in patients undergoing cardiopulmonary bypass or with major depression ([44, 45]). Animal studies showed that C4 did not affect the expression of IL-6 in mice ([46]). One possible explanation for this discrepancy could be species differences. IL-6 has recently been shown to be a critical regulator of Th17 differentiation and is seen as a marker of inflammation. Increased production of IL-6 in BD patients has been reported previously ([32, 47]). Not surprisingly, IL-6–knockout mice were shown to be resistant to the induction of experimental autoimmune uveoretinitis (EAU), an animal model of human endogenous uveitis ([48]). Additionally, IL-6 blockade showed a therapeutic effect on EAU ([49, 50]), suggesting that it plays an important role in the pathogenesis of EAU. Taken together, these data suggest that the high copy number of C4A may play a role in BD pathogenesis via enhanced expression of IL-6. We did not find an effect of C4 copy number on the activation of PBMCs and the production of other proinflammatory cytokines.

It should be noted that the present study has certain limitations that need to be taken into account when considering its contributions. The patients were recruited from our ophthalmic center and may represent a selected population of patients. Confirmation of our data is therefore needed by studies of patients recruited from other medical specialties, such as rheumatology, to confirm our data for the BD and AS patient groups. The results presented here also need to be validated in other ethnic cohorts.

In conclusion, our results indicate that a high copy number of C4A confers risk for BD, but not AS-associated acute anterior uveitis, in a Han Chinese population. A higher C4A copy number may be associated with an effect on the regulation of C4A expression and control of a proinflammatory cytokine such as IL-6. These findings highlight the important role of C4 in the pathogenesis of BD.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Yang had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Hou, Yang.

Acquisition of data. Hou, Qi, Liao, Fang.

Analysis and interpretation of data. Hou, Qi, Liao, Q. Zhang, Fang, Zhou, Liu, Bai, M. Zhang, Kijlstra, Yang.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES