Advertisement

Inflammation Research

, Volume 67, Issue 5, pp 423–433 | Cite as

HIF-1A gene polymorphisms and its protein level in patients with rheumatoid arthritis: a case–control study

  • Agnieszka Paradowska-Gorycka
  • Barbara Stypinska
  • Andrzej Pawlik
  • Ewa Haladyj
  • Katarzyna Romanowska-Próchnicka
  • Marzena Olesinska
Original Research Paper

Abstract

Objectives

The aim of the study was to identify HIF-1A genetic variants and their possible association with HIF-1α, VEGF, KDR, RORc and Foxp3 protein levels, and susceptibility to and severity of RA.

Methods

The HIF-1A gene polymorphisms were genotyped for 587 RA patients and 341 healthy individuals. The HIF-1α, VEGF, KDR, RORc and Foxp3serum levels were evaluated.

Results

Under the codominant model, the frequency of the rs12434438 GG genotype was lower in RA patients than in controls (P = 0.02). Under the recessive model (AA + AG vs GG), the association was also significant (OR 3.32; CI 1.19–9.24; P = 0.02). Overall, rs12434438 A/G and rs1951795 A/C are in almost completed linkage disequilibrium with D′ = 0.96 and r2 = 0.85. The HIF-1A rs1951795 A allele was associated with rheumatoid factor (P = 0.02) and mean value of erythrocyte sedimentation rate (ESR) (P = 0.05). In RA patients with HIF-1A rs12434439 GG genotype, the parameters of disease activity such as DAS-28, VAS score, Larsen score or HAQ score were lower compared to RA patients with the HIF-1A rs12434439 AA genotype. Moreover, we also observed that Foxp3 serum levels were higher, and RORc2 serum levels were lower in RA patients with rs12434439 GG.

Conclusion

The polymorphic HIF-1A rs12434439 GG genotype may play a protective role for RA development.

Keywords

Gene polymorphisms Rheumatoid arthritis HIF-1α Angiogenesis Inflammation 

Introduction

Rheumatoid arthritis (RA) is a common inflammatory and angiogenic disease. While the importance of inflammation in RA is well understood, little is known about the molecular mechanisms promoting angiogenesis in RA [1, 2]. However, recent studies have demonstrated that induction of both these processes is likely a consequence of synovial hypoxia, which is defined as low oxygen partial pressure [3, 4]. Synovial hypoxia of the inflamed joints is thus considered as potential pathogenic factor in RA [5], and can play a role not only in reinforcing synovial inflammation but also takes place in synovium at the pre-arthritic stage [6]. The oxygen-sensitive transcription factor, which allows for adaptation of hypoxia to joints environment in rheumatoid arthritis is hypoxia-inducible factor 1 alpha (HIF-1α). This factor is overexpressed in the synovial lining and stroma cells of RA patients [7] and participates in the pathogenesis through the induction of profound changes of the expression of key genes involved in arthritis biological processes such as angiogenesis or T cell differentiation. Regulation of HIF-1α-dependent gene expression requires hypoxia-response elements (HRE) in the promotors of target genes [2, 7].

One of the best characterized HRE-containing genes is vascular endothelial growth factor (VEGF) and its receptor VEGF-R2 (also known as KDR), whose increased levels in RA patients have been described in many studies [8, 9, 10, 11]. VEGF is associated with disease activity, inflammatory markers, destructive changes, pathological features of arthritis, and increased risk of cardiovascular disease (CVD) as well as angiogenesis in patients with RA [11]. Furthermore, VEGF-induced angiogenesis in RA may be significantly reduced by inhibition of HIF-1 expression indicating that HIF-1/VEGF axis may be the primary target of RA treatment [4].

HIF-1α, a key metabolic sensor, may also have a role in the regulation of the T‑helper‑17 cells (Th17) and regulatory T (Treg) cells differentiation [2, 12], which further highlights the potential link between inflammation and hypoxia [2]. On the one hand, HIF-1α promotes Th17 development by direct transcriptional activation of RORc. On the other, HIF-1α inhibits Tregs development through binding Foxp3 and targeting its protein for degradation [12]. Importantly, HIF-1α not only regulates Foxp3 and RORc mRNA expression, but also interacts with their proteins to regulate the expression of Treg- and Th17-related genes, respectively [12].

A several reports have demonstrated that the overexpression of HIF-1α promotes the expression of proinflammatory cytokines, growth factors and autoantibody production, thus exacerbating the severity of RA [5, 13, 14]. Because HIF-1α participates in the pathogenesis of RA mainly in two aspects, promoting angiogenesis and regulating inflammation, it may be an interesting candidate for a novel marker for RA. Moreover, single nucleotide polymorphisms (SNPs), which are attractive biomarkers for translational studies, may regulate the expression and stability of both HIF-1α mRNA and protein. Given the important role of HIF-1α in the pathogenesis of RA, we examined four functional SNPs in the HIF-1A gene and evaluated their association with the HIF-1α, VEGF, KDR, RORc and Foxp3 protein levels as well as with susceptibility to and severity of RA in a Polish population.

Materials and methods

Study subjects

A total of 587 patients with RA were recruited from the National Institute of Geriatrics, Rheumatology and Rehabilitation in Warsaw, Poland, and Pomeranian Medical University in Szczecin, Poland.

Patients were classified by rheumatologists, according to the revised 1987 American College of Rheumatology (ACR) criteria (511 patients) and 2010 EULAR/ACR criteria (76 patients). Only patients who fulfilled the above criteria were enrolled. All RA patients were assessed by (DAS-28—a disease activity score for 28 joints) which included: the number of swollen and tender joints and erythrocyte sedimentation rate (ESR); a visual analog scale (VAS). Patients’ global status was calculated using the Health Assessment Questionnaires (HAQ): range 0–3. Furthermore, levels of quantification of C-reactive protein (CRP), platelets (PLT) and creatinine (CRT) were also determined by standard methods. The presence of anti-cyclic citrullinated peptide (anti-CCP) antibodies (≥ 17 U/ml) was determined using commercial kits (Elecsys Anti-CCP assay; Roche Diagnostics GmbH, Basel, Switzerland) with measuring range 7–500 U/ml, and the presence of rheumatoid factor (RF) (≥ 34 U/ml) was determined by the nephelometric method. Radiological progression was assessed by a Larsen method. RA patients were divided according to the coexistence of hypertension, heart disease and organ involvement.

A total of 341 healthy individuals (217 females and 124 males) were randomly selected from blood bank donors. Exclusion criteria in these group were: the presence of ANA antibodies and of arthritis and any other clinical or laboratory signs of autoimmune diseases.

Patients and control subjects had the same socioeconomic status and were from the same geographical area. All subjects were of European ancestry. We selected a representative sample of the admixed urban Polish population. Informed consent was obtained from all individual participants included in the study. The study was reviewed and approved by the Research Ethics Committee of the National Institute of Geriatrics, Rheumatology and Rehabilitation, and Pomeranian Medical University. All procedures performed in this study were in accordance with the ethical standards of our institute and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.

DNA extraction and HIF-1A genotyping

Peripheral blood (PB) was collected in vacutainers containing EDTA. DNA was extracted from PB samples of all participants using the standard isothiocynate guanidine (GTC) extraction method and/or the QIAamp DNA Blood Mini Kit (Qiagen, Hilden, Germany). The concentration of the DNA obtained was about 30 ng/ul. The TaqMan allelic discrimination assay was used to detect the genotypes of rs11549465, rs12434438, rs1951795 and rs2057482 HIF1A gene polymorphisms (Applied Biosystems, Foster City, CA, USA). The reaction was performed in 10 ul volume on StepOne Real-Time PCR system in RotorGene 6000 with the fallowing amplification protocol: denaturation at 95 °C for 10 min, followed by 40 cycles of denaturation at 92 °C for 15 s, and annealing and extension at 60 °C for 1 min. Negative controls and duplicate randomly selected samples were included to check the accuracy of genotyping.

HIF-1α protein at the serum level

Commercially available enzyme-linked immunosorbent assay (ELISA) kits were used for measuring HIF-1α levels in the serum of RA patients and controls (ELISA kit HIF 1 alfa SunRed Biotechnology Company, Shanghai, China). Estimation of HIF-1α was performed in accordance with the manufacturer’s instructions. The minimum level of detection for HIF-1α was 0.946 pg/ml. The intra-assay coefficient of variation was < 10%. Plates were read at an absorbance of 450 nm on LT-4000MS reader (Labtech International Ltd, Great Britain).

Assay for serum levels of VEGF, KDR, RORc and Foxp3

Serum samples were separated from peripheral venous blood and collected at − 86 °C until analysis.

VEGF levels were determined using human VEGF ELISA kits (R&D systems, Minneapolis, MN, USA), according to the manufacturer’s instructions. The minimum level of detection for VEGF was 9 pg/ml.

KDR levels were determined using human sVEGFR2/KDR/Flk-1 ELISA kits (R&D systems, Minneapolis, MN, USA), according to the manufacturer’s instructions. The minimum level of detection for KDR was 4.6 pg/ml.

The levels of circulating Foxp3 in serum were determined using Foxp3 ELISA kits (Uscn Life Science), according to the manufacturer’s instructions. The minimum level of detection for Foxp3 was 0.121 ng/ml.

Serum RORc levels were determined by ELISA kits (Wuhan EIAab Science Co., Ltd, Wuhan, China), according to the manufacturer’s instructions. The minimum level of detection for RORc was 0.15 ng/ml.

The plates were read using an ELISA reader (El × 800, BIO-TEK Instruments, Winooski, VT, USA) at 450 nm.

Statistical analysis

Comparisons of the general characteristics between different groups were statistically evaluated using data analysis software system SAS Enterprise Guide (SAS Institute Inc., Cary, NC, USA. 2013, version 6.1 M1) and STATISTICA [StatSoft. Inc. (2011), version 10]. Non-normally distributed continuous variables were presented as median and (interquartile range—IQR), categorical variables were presented as percentages. The normality of the distribution was tested using the Shapiro–Wilk test. All polymorphisms were tested for deviation from Hardy–Weinberg equilibrium (HWE) [Michael H. Court (2005–2008)]. Linkage disequilibrium (LD) and coefficient (D′ and r2) for haplotypes and their frequencies were performed using the genetic statistical software SHEsis (http://analysis.bio-x.cn) [15, 16, 17].

The χ2 analyses or Fisher exact probability test was used to evaluate differences in genotype and allele distribution between the examined groups (unadjusted OR and 95% confidence intervals, unadjusted P value). Tests were performed under four genetic models including codominant, dominant, recessive and overdominant models. Logistic regression was used to evaluate the effect of possible confounders such as age and gender (adjusted OR, adjusted P value).

The association between SNPs and clinical/serological parameters was compared by the Mann–Whitney test and χ2 or Fisher exact test (depending on expected values) for categorical variables. Differences were considered statistically significant if the two-sided P value was  < 0.05.

The relation between SNPs and HIF-1α and VEGF protein level was evaluated using Kruskal–Wallis test. Bonferroni adjusted P value < 0.0167 was considered to be significant.

Results

Patient’s characteristics

The demographic, clinical and biochemical data were collected from patients at the time of blood sampling and are summarized in Table 1. The median age of RA patients was 56 years. All included RA patients showed long-standing disease with disease duration > 10 years, DAS-28 score > 4.99, HAQ score 1.5 and Larsen score 3. Among all 587 RA patients, 347 (69%) were rheumatoid factor (RF) positive and 250 (81%) had anti-citrullinated peptide/protein antibody (ACPA). Table 2 shows the characteristics of RA patients with and without cardiovascular events. Evidence of coronary artery disease was found in 14% of patients, hypertension in 36% and myocarditis in 3%; all these symptoms were classified as CVD. Patients with CVD events were older than those without CVD events (63.5 vs 56 years, respectively, P < 0.0001) and had a higher disease activity. There was no significant difference in disease duration between those patients.

Table 1

Distribution of RA patients’ characteristics and prognosis analysis

Characteristics

RA patients

Healthy subjects

N

Median (IQR)

N

Median (IQR)

Age [years]

587

56 (50–65)

341

28 (23–37)

Disease duration [years]

469

10 (5–15)

  

Larsen

511

3 (3–3)

  

Number of swollen joints

303

3 (1–7)

  

Number of tender joints

303

7 (3–12)

  

ESR [mm/h]

508

30 (16–48)

  

CRP [mg/L]

307

13 (6–32)

  

Hemoglobin [g/dL]

306

12.7 (11.6–13.5)

  

VAS [mm]

299

52 (31–70)

  

DAS 28-CRP

300

4.995 (3.82–5.93)

  

HAQ

286

1.5 (0.875–2)

  

PLT [x103/mm3]

306

312.5 (255–383)

  

Creatinine

305

0.7 (0.6–0.8)

  
 

N

n (%)

N

n (%)

Sex (female)

587

526 (88.11%)

341

217 (63.64%)

RF presence

506

347 (68.58)

  

Anti-CCP presence

310

250 (80.65)

  

Morning siffness

283

230 (81.27)

  

Organ symptoms

513

108 (21.05)

  

Coronary artery disease

305

42 (13.77)

  

Hypertension

306

111 (36.27)

  

Myocarditis

303

10 (3.30)

  

N number of patients with clinical information, n number of patients with positive clinical manifestation, IQR interquartile range, DAS-28 disease activity score for 28 joints, VAS visual analog scale (range 0–100), HAQ Health Assessment Questionnaires (range 0–3), CRP C-reactive protein, ESR erythrocyte sedimentation ratio, PLT platelet, RF rheumatoid factor (> 34 IU/ml), anti-CCP anti-CCP antibodies (> 17 U/ml)

Table 2

Demographic and clinical characteristics of RA patients with and without CVD

Parameter continuous variables

Patients with cardiovascular diseases

Patients without cardiovascular diseases

P

N

Median (IQR)

N

Median (IQR)

Age [years]

42

63.5 (72–59)

263

56 (64–48)

< 0.0001

Disease duration [years]

30

11 (19–7)

230

10 (17–6)

0.37

Number of swollen joints

41

4 (8–2)

258

3 (7–1)

0.27

Number of tender joints

41

8 (14–4)

258

7 (12–3)

0.31

Larsen

42

3 (4–2)

262

3 (4–3)

0.90

ESR [mm/h]

42

20 (33–12)

261

26.9 (42–13)

0.21

CRP [mg/L]

40

26 (42.1–12.85)

261

12 (30.5–5.1)

0.001

VAS [mm]

40

64 (78–44.5)

256

51 (68.5–30.5)

0.01

DAS-28

40

5.24 (6.13–4.6145)

257

4.95 (5.89–3.76)

0.03

HAQ

41

1.75 (2.125–1.375)

236

1.5 (2–0.875)

0.04

Hb

41

13 (13.5–11.7)

260

12.6 (13.5–11.5)

0.50

PLT

41

276 (343–233)

260

319 (384.5–262)

0.08

Creatinine

40

0.8 (0.9–0.69)

260

0.7 (0.8–0.6)

< 0.001

Parameter discrete variables

Patients with cardiovascular diseases

Patients without cardiovascular diseases

P

N

n (%)

N

n (%)

Women

42

36 (85.71)

263

246 (93.54)

0.1068

RF positive

41

27 (65.85)

257

167 (64.98)

0.91

Anti-CCP positive

41

33 (80.49)

261

211 (80.84)

0.96

N no. of patients with clinical information, n no. of patients with positive clinical manifestation, IQR interquartile range

Information about HIF-1A gene polymorphisms

The four HIF-1A (NC_000014.9) SNPs studied, rs11549465, rs12434438, rs1951795 and rs2057482, were selected on the basis of previous publications [18, 19, 20, 21]. SNPs with minor allele frequency (MAF) < 0.05 (< 5%) were excluded. For the analysis we identified polymorphisms with a functional activity and which are associated with other disease in previous study. Polymorphism at position rs11549465 (C1772T; Pro582Ser) is located in exon 12 of the HIF-1A gene, within the ODD/pVHL domain, and influence the transcriptional activity of this gene to increase dramatically in comparison with the common isoform [21]. The second examined polymorphism at position rs2057482 is located near the miR-199a seed-binding site in the 3′UTR of HIF-1A gene and it could lead to differential regulation of HIF-1α protein by miR-199a [22]. In contrast, SNPs at position rs12434438 and rs1951795 are located in the non-coding region of HIF-1A gene introns and their function has not been demonstrated yet. The MAF of the SNPs in our samples were similar to those in the Utah residents of northern and western European ancestry (HapMap database; Table 1S in supplementary files). Genotype distributions of all examined HIF-1A polymorphisms were in Hardy–Weinberg equilibrium (HWE) in both RA patients and control group. Also, there was no evidence of any systematic bias in genotyping.

Association of the individual SNPs with risk of rheumatoid arthritis (RA)

First to confirm the genotyping results, PCR-amplified randomly selected DNA samples were analyzed by direct sequencing, using an ABI PRISM Sequencer (Applied Biosystems, Foster City, CA, USA), and the results were 100% concordant. The genotyping success was greater than 96% in all cases. The distributions of genotype and allele frequencies of the examined HIF-1A gene polymorphisms among RA patients and healthy subjects are shown in Table 3. We calculated different genetic models such as codominant, dominant, recessive and overdominant, to assess the association of the HIF-1A genetic variants with the probability of the risk of acquiring rheumatoid arthritis in our Polish population. We found that HIF-1A rs12434438 A/G polymorphism revealed differences in the case–control distribution. Under the codominant model, the frequency of the rs12434438 GG genotype was lower in RA patients compared to the controls (P = 0.02). Similarly, under the recessive model (AA + AG vs GG), the association was also significant (OR 3.32; CI 1.19–9.24; P = 0.02). No significant differences in the proportion of cases and control under diffrent models were found for other studied HIF-1A gene polymorphisms. Effect sizes were adjusted for sex and age.

Table 3

Distribution of genotypes and allele frequencies of HIF-1α SNPs among patients with RA and healthy subjects (P = RA vs controls)

rs12434438 A/G

Genotype

RA

n (%)

Controls

n (%)

OR (95% CI)a

P*

Codominant

AA

365 (65)

213 (66)

AG

181 (32)

96 (30)

1.02 (0.64–1.62)

0.94

GG

18 (3)

15 (4)

3.37 (1.19–9.56)

0.02

Dominant

AA

365 (65)

213 (66)

GG + AG

199 (35)

111 (34)

1.17 (0.75–1.82)

0.48

Recessive

AA + AG

546 (97)

309 (95)

GG

18 (3)

15 (5)

3.32 (1.19–9.24)

0.02

Overdominant

AA + GG

383 (68)

228 (70)

AG

181 (32)

96 (30)

0.95 (0.60–1.49)

0.82

rs1951795 A/C

 

RA

n (%)

Controls

n (%)

OR (95% CI)a

P*

Codominant

CC

396 (68)

224 (67)

AC

172 (29)

102 (30)

1.27 (0.80–2.00)

0.31

AA

19 (3)

11 (3)

1.88 (0.55–6.46)

0.32

Dominant

CC

396 (67)

224 (66)

AC + AA

191 (33)

113 (34)

1.3 (0.84–2.03)

0.23

Recessive

CC + AC

568 (97)

326 (97)

AA

19 (3)

11 (3)

1.72 (0.52–5.72)

0.37

Overdominant

AA + CC

415 (71)

235 (70)

AC

172 (29)

102 (30)

1.23 (0.79–1.93)

0.36

rs2057482 C/T

 

RA

n (%)

Controls

n (%)

OR (95% CI)a

P*

Codominant

CC

449 (78)

254 (77)

CT

6

2

1.55 (0.94–2.55)

0.09

TT

123

75

0.21 (0.03–1.64)

0.14

Dominant

CC

449 (78)

254 (77)

CT + TT

129 (22)

77 (23)

1.41 (0.87–2.31)

0.17

Recessive

TT

572 (99)

329 (99)

CC + CT

6 (1)

2 (1)

0.19 (0.02–1.50)

0.11

Overdominant

CT

455 (79)

256 (77)

CC + TT

123 (21)

75 (23)

1.58 (0.96–2.60)

0.07

rs11549465 C/T

 

RA

n (%)

Controls

n (%)

OR (95% CI)a

P*

Codominant

CC

485 (85)

283 (86)

CT

85 (15)

43 (13)

1.14 (0.64–2.03)

0.66

TT

0

1 (1)

0.99

Dominant

CC

485 (85)

283 (86)

CT + TT

85 (15)

44 (14)

1.16 (0.65–2.06)

0.62

Recessive

CC + CT

570 (100)

326 (99.69)

TT

0 (0.00)

1 (0.31)

0.99

Overdominant

CC + TT

485 (85)

284 (87)

CT

85 (15)

43 (13)

1.13 (0.63–2.03)

0.67

Statistically significant values are in bold

P* χ2 test with Yates’ correction

p = RA vs controls

p ≤ 0.05 was considered as significant

aAdjusted

Linkage analysis and haplotypes association with the risk of RA

To further confirm the contribution of HIF-1A gene polymorphisms with RA susceptibility, we enforced the haplotypes association analysis. We genetarted a Haploview plot on the 4 HIF-1A gene polymorphisms (Fig. 1). Analysis revealed very high linkage disequilibrium (LD) between rs12434438 A/G and rs1951795 A/C with D′ = 0.96 and r2 = 0.85. High D′ was also observed between other examined polymorphisms—D′ > 0.86, but because r2 was < 0.49 the SNPs cannot substitute each other and get all possible existing haplotypes.

Fig. 1

Halploview plot of HIF-1A gene polymorphisms. The plot illustrates pairwise linkage disequilibrium (LD) between the examined SNPs based on the D′ and r2 values. Values approaching zero indicate the absence of LD, and those approaching 100 indicate complete LD. The red-colored square represents varying degrees of LD < 1 and LOD (logarithm of odds) > 2 scores (strong LD) and white blocks represent varying degrees of LD < 1 and < 2 scores (weak LD). site1 rs12434438 A/G, site 2 rs1951795 A/C, site 3 rs2057482 C/T, site 4 rs11549465 C/T. (Color figure online)

To create the haplotypes, HIF-1A genetic variants were analyzed in the following sequence: rs12434438 A/G, rs1951795 A/C, rs2057482 C/T, rs11549465 C/T. 12 potential haplotypes were formed; however, the haplotypes with frequency greather than 3% are presented (Table 4) and were further analyzed. Among them, the most frequent haplotype was ACCC with a frequency of 77% in RA patients and 80% in healthy subjects.In contrast, the least frequent haplotype identified in both RA patients and controls was the GATC haplotype (with frequency about 5%). However, there were no differences in haplotype frequency between RA patients and controls for all presented haplotypes.

Table 4

HIF-1 haplotypes in RA patients and controls

Haplotype rs12434438 A/G rs1951795 A/C rs2057482 C/T rs11549465 C/T

RA

n = 1051 (%)

Control

n = 589 (%)

OR (95% CI)

P

ACCC

861 (77.2)

482 (80.3)

0.833 (0.651–1.064)

0.157

GATT

73 (6.5)

40 (6.6)

0.981 (0.658–1.463)

0.919

GATC

53 (4.7)

30 (5)

0.949 (0.599–1.502)

0.814

GACC

64 (5.7)

37 (6.1)

0.927 (0.61–1.408)

0.747

P Fisher’s test; p considered as significant is in bold

Correlation between HIF-1A gene polymorphisms and rheumatoid arthritis phenotype

Although we found a lack of correlation between HIF-1A gene polymorphisms and susceptibility to RA, in the next step we investigated the potential association between all examined polymorphisms and RA phenotype according to clinical/laboratory characteristics. We performed a stratified analysis of combined genotypes under the dominant and the recessive models for each examined polymorphism. Our data showed that the analyzed HIF-1A gene polymorphisms do not have a high impact on RA phenotype. The genotype–phenotype analysis showed correlation of the HIF-1A rs1951795 A/C polymorphism with rheumatoid factor (RF) presence (P = 0.02) under the dominant model (CC vs CA + AA; Table 2S in supplementary files) as well as with the mean value of erythrocyte sedimentation rate (ESR) (P = 0.05) and PLT (P = 0.06) under the recessive model (AA vs CA + CC; Table 3S in supplementary files). In our study, we also observed that in carriers of the rs12434438 AA genotype and rs2057482 CC genotype, there was a tendency to have more frequent RF presence than in patients with other genotypes (P = 0.09 and P = 0.06, respectively; Table 4S and 5S in supplementary files). Moreover, we also observed that carriers of the combined rs11549465 CT + TT genotype had a higher median value of ESR in comparison to rs11549465 CC subjects (P = 0.06, Table 6S in supplementary files).

HIF-1α serum levels in relation to disease activity

Given the important role of HIF-1α on the production of proinflammatory cytokines as well as Th17/Treg balance, we tasted whether serum levels of HIF-1α may have an impact on the clinical activity of RA patients. As shown in Table 5, patients with RA were divided into two groups: I group included patients with higher disease activity as well as RF, ACPA and CVD positive, whereas II group included RA patients with the lowest clinical activity of disease and without RF, ACPA and CV disease. We observed that the median value of HIF-1α serum levels was higher in RA patients with disease duration > 10 years (P = 0.003).

Table 5

Clinical characteristics of RA patients in relation to HIF-1α serum levels

Parameter

 

HIF protein level

 

HIF protein level

P

Parameter group I

N

Median (IQR)

Parameter group II

N

Median (IQR)

Age

Age > 56

287

88.19 (149.73–58.50)

Age ≤ 56

314

112.03 (161.33–60.32)

0.22

Sex

Women

526

102.59 (157.71–59.45)

Men

71

79.53 (146.2–58.25)

0.25

RF

RF+

347

100.52 (156.66–62.42)

RF−

159

102.98 (157.51–59.95)

0.66

anti-CCP

a-CCP+

250

93.61 (156.66–60.152)

a-CCP−

60

133.73 (158.4–69.9)

0.23

Disease duration

≥ 10

245

127.21 (166.00–79.53)

< 10

356

79.37 (149.33–57.30)

0.003

ESR

≥ 30

258

112.48 (163.27–59.95)

< 30

343

95.47 (151.8–58.66)

0.49

Number of tender joints

≥ 7

164

102.58 (162.64–65.06)

< 7

437

89.52 (151.8–55.92)

0.15

Number of swollen joints

≥ 3

169

92.23 (155.4–58.74)

< 3

432

115.22 (157.99–59.03)

0.64

CRP

≥ 13

158

95.41 (151.98–62.506)

< 13

443

107.32 (158.52–56.53)

0.91

DAS-28

≥ 5.0

150

91.17 (156.66–59.31)

< 5.0

451

111.56 (157.51–59.02)

0.80

HAQ

≥ 1.5

156

87.24 (147.78–61.25)

< 1.5

445

114.18 (158.64–58.66)

0.47

Cardiovascular diseases

CAD+

42

78.47 (155.4–54.06)

CAD−

263

104.59 (157.91–64.59)

0.26

Impact of HIF-1A genetic variants on HIF-1α serum levels

First, we compared thef HIF-1α serum levels between RA patients and healthy subjects according to all examined HIF-1A gene SNPs (Table 6). The serum levels of HIF-1α in healthy subjects with rs2057482 CC, rs11549465 CC and rs1951795 CC genotypes were significantly higher than those of RA patients with the same HIF-1A genotypes (P = 0.004, P = 0.01 and P = 0.01, respectively). The tendency to have increased HIF-1α serum levels was also observed in controls with rs12434438 AA genotype compered to patients with the same genotypes (P = 0.08).

Table 6

Variation in HIF-1α serum levels in RA patients and controls in relation to HIF-1A genotypes

Genotype

HIF protein level [pg/ml] RA group

HIF protein level [pg/ml] control group

P

N

Median (IQR)

N

Median (IQR)

rs12434438 A/G

 AA

365

107.32 (153.13–60.87)

213

110.94 (161.66–76.12)

0.08

 AG

181

87.09 (159.45–57.02)

96

104.74 (160.47–65.98)

0.16

 GG

18

140.84 (197.29–85.96)

15

99.66 (140.31–77.93)

0.31

rs1951795 C/A

 CC

396

94.98 (151.7–59.03)

224

114.07 (165.39–76.06)

0.015

 CA

172

92.85 (158.4–57.14)

102

104.28 (156.86–65.96)

0.27

 AA

19

142.86 (163.96–130.76)

11

106.97 (145.68–78.38)

0.25

rs2057482 C/T

 CC

449

95.41 (151.7–58.34)

254

111.9 (165.39–76.12)

0.004

 CT

123

124.43 (170.69–60.32)

75

101.88 (145.68–66.05)

0.79

 TT

6

86.96 (230.63–84.97)

2

122.89 (167.4–78.38)

0.85

rs11549465 C/T

 CC

485

97.82 (151.7–58.66)

283

115.2 (165.39–76.116)

0.01

 CT

85

116.80 (165.25–59.95)

43

88.70 (141.34–61.22)

0.68

 TT

0

1

78.38 (78.38–78.38)

P < 0.05 was considered significant, P values in boldface are considered significant

IQR interquartile range, P Mann–Whitney U test

Second, we carried out an analysis of HIF-1α serum levels in RA patients in relation to all studied HIF-1A genotypes (Table 7S in supplementary files). However, in this case, we found no significant correlation between HIF-1A genotypes and its serum levels among RA patients.

Table 7

Impact of HIF-1A polymorphisms on VEGF, KDR, RORc and Foxp3 serum levels

 

rs12434439

P

AA

AG

GG

N

Mean ± SD or median (IQR)

N

Mean ± SD or median (IQR)

N

Mean ± SD or median (IQR)

VEGF

369

306.61 (513.69–159.27)

178

297.43 (498.34–166.59)

21

288.15 (432.55–147.44)

0.82

KDR

337

7715.18 ± 2958.19

158

7931.52 ± 2769.95

21

6994.71 ± 2039.53

0.34

RORc

341

2.04 (3.36–1.27)

161

1.95 (3.26–1.43)

21

1.76 (3.47–1.33)

0.90

Foxp3

358

0.05 (0.66–0.02)

173

0.04 (0.52–0.02)

28

0.46 (2.22–0.04)

0.06

 

rs1951795

P

CC

CA

AA

N

Mean ± SD or median (IQR)

N

Mean ± SD or median (IQR)

N

Mean ± SD or median (IQR)

VEGF

404

313.27 (513.58–162.03)

179

305.11 (508.52–166.59)

21

295.32 (653.98–166)

0.86

KDR

364

7738.77 ± 2924.54

163

7974.57 ± 2763.06

21

7007.93 ± 1841.69

0.30

RORc

369

2.05 (3.37–1.27)

168

1.96 (3.34–1.41)

21

1.8 (3.47–1.33)

0.94

Foxp3

387

0.05 (0.64–0.02)

181

0.05 (0.54–0.03)

26

0.4 (1.80–0.04)

0.07

 

rs2057482

P

CC

CT

TT

N

Mean ± SD or median (IQR)

N

Mean ± SD or median (IQR)

N

Mean ± SD or median (IQR)

VEGF

462

316.34 (514.97–162.57)

125

288.15 (509.78–158.23)

7

418.36 (495.54–206.68)

0.79

KDR

417

7780.61 ± 2904.31

114

7743.83 ± 2750.36

7

7468.32 ± 1066.07

0.95

RORc

423

2.03 (3.37–1.27)

118

1.89 (3.47–1.34)

7

2.23 (4.88–1.41)

0.87

Foxp3

446

0.05 (0.65–0.03)

131

0.06 (0.53–0.02)

7

1.43 (6.56–0.04)

0.11

 

rs11549465

P

CC

CT

TT

N

Median (IQR)

N

Median (IQR)

N

Median (IQR)

VEGF

504

314.79 (513.87–162.4)

78

290.06 (518.98–166.59)

0

0.32

KDR

453

7799.78 ± 2871.09

73

7822.30 ± 2880.89

1

6472.68

0.9

RORc

460

2.07 (3.39–1.29)

84

0.06 (0.51–0.03)

1

0.04

0.86

Foxp3

487

0.05 (0.77–0.03)

84

0.06 (0.51–0.03)

1

0.04

0.86

Impact of HIF-1A gene polymorphisms on VEGF, KDR, RORc2 and Foxp3 protein expression

Finally, we investigated whether HIF-1A SNPs had an impact on VEGF, KDR (VEGFR2), Foxp3 and RORc2 levels in serum. To assess the true contribution of the examined SNPs on the VEGF, KDR, RORc2 and Foxp3 protein levels, both study groups—RA patients and controls, where mixed together. This approach allows us looking at the effect of the polymorphisms in all subjects. As shown in Table 7, we did not observe significant correlation between HIF-1A different genotypes and VEGF, KDR, RORc2 and Foxp3 levels in serum. However, we observed a tendency to have higher Foxp3 expression in serum in subjects with rs12434439 GG (P = 0.06) and rs1951795 AA (P = 0.07) genotypes.

Discussion

In the present study, for the first time, we analyzed whether HIF-1A gene polymorphisms are associated with susceptibility to and severity of RA patients. Because, HIF-1A gene regulates the expression of over 100 genes which play an important role in controlling major cellular function such as cytokine production, angiogenesis or cell differentiation. We decided to test, for the first time, whether genetic variants located in the HIF-1A gene may have an impact on the VEGF, KDR, RORc and Foxp3 serum levels.

Limited data are available on HIF-1A genetic variants effect on arthritis processes in the different populations. In contrast, several studies have demonstrated that HIF-1A SNPs were implicated in the risk of common human diseases such as prostate cancer [19], pancreatic ductal adenocarcinoma [22], chronic obstructive pulmonary disease [23, 24], type 2 diabetic nephropathy [25], lung cancer [26] and hepatocellular carcinoma [27]. In this study, we found that the polymorphic HIF-1A rs12434439 GG genotype may play a protective role for RA development in Polish subjects. Furthermore, we demonstrated no correlation between the HIF-1A genetic variants and phenotype of RA as well as cardiovascular events in our RA patients. In contrast, López-Reyes et al. [20] explored three polymorphisms in HIF-1A (rs11549465, rs11549467, rs2057482) gene in coronary artery disease (CAD) and concluded that only HIF-1A rs2057482 SNP is involved in the risk of developing CAD and is associated with some metabolic parameters and cardiovascular risk factors. They found that the rs2057482 T allele was associated with decreased risk of obesity, hypertension, hypertriglyceridemia, hypercholesterolemia as well as increased risk of T2DM. Fernández-Torres et al. [21] suggested that the HIF1A rs11549465 variant may play a protective role in the loss of articular cartilage. They concluded that this may be explained by the fact that the presence of this HIF-1A SNP confers greater stability to the HIF-1α protein.

As we know, HIF-1α is involved in the pathogenesis of rheumatoid arthritis by multiple-step pathways, and activation of the signaling cascade of this factor leads to changes in gene expression [1, 28]. These changes trigger transcription activity in several cells, leading to the expression of genes involved in cartilage destruction, angiogenesis or recruitment of T and B cells in the rheumatoid synovium [28]. HIF-1α, through the regulation of the VEGF and its receptor gene expression, favors infiltration of inflammatory cells and induction of inflammatory mediators, perpetuating synovitis in RA [2]. In the present study, we found that RA patients with HIF-1A rs12434439 GG genotype had lower VEGF and KDR serum levels than other RA patients. Moreover, VEGF is a perfect candidate for the monitoring of erosive progression, disease activity and treatment response in patients with RA. In our RA patients with HIF-1A rs12434439 GG genotype, the parameters of disease activity such as DAS-28, VAS score, Larsen score or HAQ score were lower compared to RA patients with HIF-1A rs12434439 AA genotype (although this association was not significant). The above observation may suggest that in RA patients with HIF-1A rs12434439 GG genotype, angiogenesis is limited and leads to attenuation of severity of the arthritis. Angiogenesis, which is involved in the regulation of several soluble and cell surface-bound factors, plays a central role in RA pathogenesis for a long time [11].

HIF-1α is also one of the major players in inflammation and regulates the balance between Treg and Th17 cell plasticity. In our study, we observed that Foxp3 serum levels were higher, and, at the same time, RORc2 serum levels were lower in RA patients with the HIF-1A rs12434439 GG genotype, reflecting the ongoing inflammatory process in RA patients and the attempts to keep it under control. The role of HIF-1α in Th17 as well as Treg cells differentiation was the focus of some studies; however, the results are conflicting. Wu et al. [29] have shown that in CD4 + T cells under anoxic conditions, the increasing HIF-1α expression levels promote Foxp3 expression as well as stimulate conversion of the CD4 + T cells to Treg cells. Moreover, Clambey et al. [30] demonstrated that hypoxia promotes Treg cell function and FoxP3 expression through “at least three distinct mechanisms: (1) direct transcriptional activation of FoxP3 mRNA by HIF-1α early after hypoxia, (2) a hypoxia-driven, TGF-β-dependent process required for full hypoxic induction of FoxP3 mRNA and protein following T-cell activation and differentiation, and (3) a Treg-intrinsic role for HIF-1α that is required for optimal suppressive function of Tregs in vitro and in vivo”. In contrast, Dang et al. [12] as well as Shi et al. [31] indicated that, on the one hand, HIF-1α through direct induction of RORc2 transcription promotes Th17 cell differentiation, but on the other, HIF-1α by targeting Foxp3 protein degradation inhibits Treg cell differentiation. In our opinion, discrepancies between reports may be explained by the influence of the polymorphisms located in the HIF-1A gene; however, these observation need to be clarified in future studies. Furthermore, since the expression of Foxp3 is essential for the anti-inflammatory ability of Treg cells and important for the modulation of active disease and suppression of autoimmune responses, it may be used as a parameter of disease activity [32].

In conclusion, our study provides evidence that the polymorphic HIF-1A rs12434439 GG genotype may play a protective role for RA development. We also observed some correlation between HIF-1A genetic variant and serum levels of markers of angiogenesis and/or inflammation. Further studies are needed to shed some light on this comment and offer potential new targets for the development of new personalized medicine strategy.

Notes

Acknowledgements

The technical assistance of Wieslawa Frankowska and Teresa Golaszewska is gratefully acknowledged. We are also grateful to all the RA patients and healthy subjects whose cooperation made this study possible.

Author contributions

APG conceived, designed and performed experiments, analyzed and interpreted data and wrote the first draft of the manuscript. BS performed experiments and statistical analysis. AP contributed to drafting the manuscript. KRP, EH and MO were involved in the classification of patients with rheumatoid arthritis, clinical check of patients and treatment control. All authors read and approved the final version.

Funding

The work was supported by Grant S/13, National Institute of Geriatrics, Rheumatology and Rehabilitation.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.

Supplementary material

11_2018_1134_MOESM1_ESM.doc (192 kb)
Supplementary material 1 (DOC 192 KB)

References

  1. 1.
    Wang CH, Yao H, Chen LN, Jia JF, Wang L, Dai JY, et al. CD147 induces angiogenesis through a vascular endothelial growth factor and hypoxia-inducible transcription factor 1α-mediated pathway in rheumatoid arthritis. Arthritis Rheum. 2012;64:1818–27.CrossRefPubMedGoogle Scholar
  2. 2.
    Konisti S, Kiriakidis S, Paleolog EM. Hypoxia—a key regulator of angiogenesis and inflammation in rheumatoid arthritis. Nat Rev Rheumatol. 2012;8:153–62.CrossRefPubMedGoogle Scholar
  3. 3.
    Quiñonez-Flores CM, González-Chávez SA, Pacheco-Tena C. Hypoxia and its implications in rheumatoid arthritis. J Biomed Sci. 2016;23:62.CrossRefPubMedPubMedCentralGoogle Scholar
  4. 4.
    Deng W, Feng X, Li X, Wang D, Sun L. Hypoxia-inducible factor 1 in autoimmune diseases. Cell Immunol. 2016;303:7–15.CrossRefPubMedGoogle Scholar
  5. 5.
    Hu F, Liu H, Xu L, Li Y, Liu X, Shi L, et al. Hypoxia-inducible factor-1α perpetuates synovial fibroblast interactions with T cells and B cells in rheumatoid arthritis. Eur J Immunol. 2016;46:742–51.CrossRefPubMedGoogle Scholar
  6. 6.
    Jeon CH, Ahn JK, Chai JY, Kim HJ, Bae EK, Park SH, et al. Hypoxia appears at pre-arthritic stage and shows co-localization with early synovial inflammation in collagen induced arthritis. Clin Exp Rheumatol. 2008;26:646–8.PubMedGoogle Scholar
  7. 7.
    Maruotti N, Cantatore FP, Crivellato E, Vacca A, Ribatti D. Angiogeneis in rheumatoid arthritis. Histol Histopathol. 2006;21:557–66.PubMedGoogle Scholar
  8. 8.
    Koch AE, Harlow LA, Haines GK, Amento EP, Unemori EN, Wong WL, et al. Vascular endothelial growth factor. A cytokine modulating endothelial function in rheumatoid arthritis. J Immunol. 1994;152:4149–56.PubMedGoogle Scholar
  9. 9.
    Ballara S, Taylor PC, Reusch P, Marmé D, Feldmann M, Maini RN, et al. Raised serum vascular endothelial growth factor levels are associated with destructive change in inflammatory arthritis. Arthritis Rheum. 2001;44:2055–64.CrossRefPubMedGoogle Scholar
  10. 10.
    Kurosaka D, Hirai K, Nishioka M, Miyamoto Y, Yoshida K, Noda K, et al. Clinical significance of serum levels of vascular endothelial growth factor, angiopoietin-1, and angiopoietin-2 in patients with rheumatoid arthritis. J Rheumatol. 2010;37:1121–8.CrossRefPubMedGoogle Scholar
  11. 11.
    Paradowska-Gorycka A, Pawlik A, Romanowska-Prochnicka K, Haladyj E, Malinowski D, Stypinska B, et al. Relationship between VEGF gene polymorphisms and serum VEGF protein levels in patients with rheumatoid arthritis. PLoS One. 2016;11:e0160769.CrossRefPubMedPubMedCentralGoogle Scholar
  12. 12.
    Dang EV, Barbi J, Yang HY, Jinasena D, Yu H, Zheng Y, et al. Control of T(H)17/T(reg) balance by hypoxia-inducible factor 1. Cell. 2011;146:772–84.CrossRefPubMedPubMedCentralGoogle Scholar
  13. 13.
    Larsen H, Muz B, Khong TL, Feldmann M, Paleolog EM. Differential effects of Th1 versus Th2 cytokines in combination with hypoxia on HIFs and angiogenesis in RA. Arthritis Res Ther. 2012;14:R180.CrossRefPubMedPubMedCentralGoogle Scholar
  14. 14.
    Hot A, Zrioual S, Lenief V, Miossec P. IL-17 and tumour necrosis factor α combination induces a HIF-1α-dependent invasive phenotype in synoviocytes. Ann Rheum Dis. 2012;71:1393–401.CrossRefPubMedGoogle Scholar
  15. 15.
    Shen J, Li Z, Chen J, Song Z, Zhou Z, Shi Y. SHEsisPlus, a toolset for genetic studies on polyploid species. Sci Rep. 2016;6:24095.CrossRefPubMedPubMedCentralGoogle Scholar
  16. 16.
    Shi Y, He L. SHEsis, a powerful software platform for analyses of linkage disequilibrium, haplotype construction, and genetic association at polymorphism loci. Cell Res. 2005;15:97–8.CrossRefPubMedGoogle Scholar
  17. 17.
    Li Z, Zhang Z, He Z, Tang W, Li T, Zeng Z, et al. A partition-ligation-combination-subdivision EM algorithm for haplotype inference with multiallelic markers. Cell Res. 2009;19:519–23.CrossRefPubMedGoogle Scholar
  18. 18.
    Wipff J, Dieude P, Avouac J, Tiev K, Hachulla E, Granel B, et al. Association of hypoxia-inducible factor 1A (HIF1A) gene polymorphisms with systemic sclerosis in a French European Caucasian population. Scand J Rheumatol. 2009;38:291–4.CrossRefPubMedGoogle Scholar
  19. 19.
    Li P, Cao Q, Shao PF, Cai HZ, Zhou H, Chen JW. at al. Genetic polymorphisms in HIF1A are associated with prostate cancer risk in a Chinese population. Asian J Androl. 2012;14:864–9.CrossRefPubMedPubMedCentralGoogle Scholar
  20. 20.
    López-Reyes A, Rodríguez-Pérez JM, Fernández-Torres J, Martínez-Rodríguez N, Pérez-Hernández N, Fuentes-Gómez AJ, et al. The HIF1A rs2057482 polymorphism is associated with risk of developing premature coronary artery disease and with some metabolic and cardiovascular risk factors. The genetics of atherosclerotic disease (GEA) Mexican study. Exp Mol Pathol. 2014;96:405–10.CrossRefPubMedGoogle Scholar
  21. 21.
    Fernández-Torres J, Hernández-Díaz C, Espinosa-Morales R, Camacho-Galindo J, Galindo-Sevilla Ndel C, López-Macay Á, et al. Polymorphic variation of hypoxia inducible factor-1 A (HIF1A) gene might contribute to the development of knee osteoarthritis: a pilot study. BMC Musculoskelet Disord. 2015;16:218.CrossRefPubMedPubMedCentralGoogle Scholar
  22. 22.
    Wang X, Ren H, Zhao T, Ma W, Dong J, Zhang S, et al. Single nucleotide polymorphism in the microRNA-199a binding site of HIF1A gene is associated with pancreatic ductal adenocarcinoma risk and worse clinical outcomes. Oncotarget. 2016;7:13717–29.PubMedPubMedCentralGoogle Scholar
  23. 23.
    Yu ZG, Wang BZ, Cheng ZZ. The association of genetic polymorphisms of hypoxia inducible factor-1alpha and vascular endothelial growth factor with increased risk of chronic obstructive pulmonary disease: a case-control study Kaohsiung. J Med Sci. 2017;33:433–441.Google Scholar
  24. 24.
    Wei WT, Li B, Chen M, Jia HR, Zhang HX. Associations between HIF-1α polymorphisms C1772T and G1790A and susceptibility to chronic obstructive pulmonary disease. Genet Mol Res. 2015;14:17341–7.CrossRefPubMedGoogle Scholar
  25. 25.
    Bi YX, Yu L, Jin GX. Correlation between polymorphisms of hypoxia-inducible factor-1α Pro582Ser and type 2 diabetic nephropathy. Genet Mol Res. 2015;14:14503–9.CrossRefPubMedGoogle Scholar
  26. 26.
    Liao SH, Liu WZ, Liu T, Sun Y, Feng X. Zhou HF.Potential signaling pathway of hypoxia-inducible factor in lung cancer and its gene polymorphism with lung cancer risk. J Recept Signal Transduct Res. 2015;35(4):233–7.CrossRefPubMedGoogle Scholar
  27. 27.
    Guo X, Li D, Chen Y, An J, Wang K, Xu Z, et al. SNP rs2057482 in HIF1A gene predicts clinical outcome of aggressive hepatocellular carcinoma patients after surgery. Sci Rep. 2015;5:11846.CrossRefPubMedPubMedCentralGoogle Scholar
  28. 28.
    Hu F, Shi L, Mu R, Zhu J, Li Y, Ma X, et al. Hypoxia-inducible factor-1α and interleukin 33 form a regulatory circuit to perpetuate the inflammation in rheumatoid arthritis. PLoS One. 2013;8:e72650.CrossRefPubMedPubMedCentralGoogle Scholar
  29. 29.
    Wu J, Cui H, Zhu Z, Wang L, Li H, Wang D. Effect of HIF1α on Foxp3 expression in CD4+ CD25− T lymphocytes. Microbiol Immunol. 2014;58:409–415.CrossRefPubMedGoogle Scholar
  30. 30.
    Clambey ET, McNamee EN, Westrich JA, Glover LE, Campbell EL, Jedlicka P, et al. Hypoxia-inducible factor-1 alpha-dependent induction of FoxP3 drives regulatory T-cell abundance and function during inflammatory hypoxia of the mucosa. Proc Natl Acad Sci USA. 2012;109:E2784–2793.CrossRefPubMedPubMedCentralGoogle Scholar
  31. 31.
    hi LZ, Wang R, Huang G, Vogel P, Neale G, Green DR, et al. HIF1alpha-dependent glycolytic pathway orchestrates a metabolic checkpoint for the differentiation of TH17 and Treg cells. J Exp Med. 2011;208:1367–1376.CrossRefGoogle Scholar
  32. 32.
    Kao JK, Hsue YT, Lin CY. Role of new population of peripheral CD11c(+) CD8(+) T cells and CD4(+) CD25(+) regulatory T cells during acute and remission stages in rheumatoid arthritis patients. J Microbiol Immunol Infect. 2007;40:419–27.PubMedGoogle Scholar

Copyright information

© Springer International Publishing AG, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Department of Biochemistry and Molecular BiologyNational Institute of Geriatrics, Rheumatology and RehabilitationWarsawPoland
  2. 2.Department of PhysiologyPomeranian Medical UniversitySzczecinPoland
  3. 3.Department of General and Experimental PathologyWarsaw Medical UniversityWarsawPoland
  4. 4.Department of Connective Tissue DiseasesNational Institute of Geriatrics, Rheumatology and RehabilitationWarsawPoland

Personalised recommendations