R s p o 1 ( p g /m L )
A
0 10 20 30 40 0
2000 4000 6000
B
ESR (mm/Hr)
D kk 1 ( p g /m L )
0 10 20 30 40 0
10 20 30 40
C
ESR (mm/Hr)
D k k1 /R s po 1
Figure2.Serum levelsofRspo-1orDkk-1andDkk-1/Rspo-1ratios in patients with rheumatoid arthritis (RA)orankylosing spondylitis (AS)and healthy controls(HC).RA patientshad significantly lower Rspo1 levels (p<0.0001 by Kruskal-Wallis with Dunn's multiple comparison test,A),higher Dkk1 levels (p<0.0001,B),and higher Dkk1/Rspo1 ratios (p<0.0001,C) than healthy controls,while AS patientshassignificantly lowerDkk1levels(p<0.001,B).Dkk1/Rspo1 ratios tended to be lower in AS patients than those in healthy controls.Circles presentindividualsubjects.Lines show the median withinterquartilerange.NS =notsignificant
HC (n=53) RA (n=100) AS (n=60)
1 10 100 1000 10000
NS
p<0.05 p<0.05
A
Rspo1 (pg/mL)
p<0.0001
HC (n=53) RA (n=100) AS (n=60) 0
2000 4000 6000 8000
p<0.05
p<0.05 p<0.05
B
Dkk1 (pg/mL)
p<0.0001
HC (n=53) RA (n=100) AS (n=60)
0.01 0.1 1 10 100 1000
NS
p<0.05 p<0.05 C
Dkk1/Rspo1
p<0.0001
Figure 3. Comparison of serum Rspo1 levels and Dkk1/Rspo1 between erosive RA and non-erosive RA.Patients with erosive RA had significantly lowerRspo1levels(p=0.022by Mann-Whitney test. A) and higher Dkk1/Rspo1 levels(p=0.026,B) than the one with non-erosive RA.Circles presentindividualsubjects.Lines show the medianwithinterquartilerange.
Figure 4.Comparison between RA patients with high Dkk1/Rspo1 and low Dkk1/Rspo1 ratios.High Dkk1/Rspo1 ratio was defined as abovethe75percentilevalueofhealthycontrolsandlow Dkk1/Rspo1 ratiowasdefinedasbelow the25percentilevalueofhealthycontrols. ESR (p=0.025by Mann-Whitney test,A),CRP (p=0.024,B),DAS28 (p=0.025, C), and Sharp scores (p=0.022, D) were significantly increased in RA patientswith high Dkk1/Rspo1compared totheone with low Dkk1/Rspo1ratio.Circlespresentindividualsubjects.Lines show themedian.
Figure 5. Comparison between AS patients with and without syndesmophyte.AS patientwith syndesmophyteshowed significantly higher Dkk/Rspo1 ratios compared to those withoutsyndesmophyte (p=0.003byMann-Whitneytest).
Figure 6.Comparison ofserum Rspo-1 levelsorDkk1/Rspo1 ratios before and 12 weeks after anti-TNF-α therapy.In RA patients, DAS28 was significantly decreased 12 weeks after anti-TNF-α therapy (n=15,p<0.0001by Wilcoxon signed rank test,A).Likewise, the levels of serum Rspo1 tended to increase (p=0.055, B) and Dkk1/Rspo1 ratios were significantly decreased (p<0.0001, D). However,no change in serum Dkk1 levels was observed (C).After anti-TNF-α therapy for12weeks,AS patientsrevealed asignificant improvementin disease activity,ASDAS (n=10,p=0.002,E).While serum Rspo1levelsweresignificantly increased(p=0.020,F),they did notshow anychangeinserum Dkk1(G)andDkk1/Rspo1(H)inAS patients.NS =notsignificant.
Figure 7.Correlation between synovialfluid (SF)Rspo1 levels and serum CRP orbetween SF Dkk1and SF whiteblood cellcounts.In RA patient,SF levelsofRspo1wereinversely correlatedwith serum C-reactive protein (CRP) levels (Spearman correlation coefficient r=-0.663,p=0.004,A).On the other hand,SF Dkk1 levels were positively related with SF white blood cellcounts (r=0.551,p=0.022, B).
0 100 200 300 400
0 2 4 6 8 10
A
SF Rspo1 (pg/mL)
CRP (mg/dL)
0 20 40 60 80 100 0
20000 40000 60000
B
SF Dkk1 (pg/mL) WBC (/mm3 )
Figure 8.Comparison ofsynovialfluid (SF)Rspo1 levels between patientswithandwithoutosteophytesintheaspiratedjoint.Whenall SF sampleswereanalyzed together,patientswith osteophytesin the kneejointshowed atendency tohaveahigherSF Rspo1levelthan thosewithoutosteophytes.
Abst r act
Introduction:Dickkopf-1 (Dkk1),an inhibitor ofWntsignaling,is known to be involved in inflammatory jointdestruction.R-spondins (Rspo) are a family of secretory proteins that can enhance Wnt signaling by interfering with Dkk1-mediated LDL receptor-related protein 6and Kremen association.WhileRspo1protectsagainstjoint destructioninamurinearthritismodel,thecirculating levelsofRspo1 havenotbeen investigatedin patientswith rheumatoidarthritis(RA) and ankylosing spondylitis (AS).Wedetermined theserum levelsof Rspo1aswellasDkk1andstudiedclinicalimplicationsofeach level and Dkk1/Rspo1 ratio, a measure of Wnt signal suppression, in patientswithRA orAS.
Methods:Serawerecollectedfrom 100RA patients,60AS patients, and 3:1 age-and gender-matched 53 healthy controls.Twenty five paired samples (15 RA and 10 AS) were obtained from patients before and 12 weeks after treatment with anti-TNF-α agents. Additionally, we analyzed synovial fluid (SF) samples from 24 patients with knee jointeffusion (17 RA and 7 osteoarthritis (OA)). Clinicaland laboratory data included age,gender,body mass index, disease duration, current medications, complete blood count, erythrocyte sediment rate (ESR), and C-reactive protein (CRP). Diseaseactivitiesin RA and AS weremeasured by DiseaseActivity Scorein 28joints(DAS28)orankylosing spondylitisdiseaseactivity score(ASDAS),respectively.Radiographicjointdamagewasassessed
by the modified Sharp/Van der Heijde score (Sharp score)in RA.
The levels of Rspo-1 and Dkk1 were determined using sandwich ELISA.
Results: Serum Rspo1levels were significantly decreased in RA patientscomparedtoothergroups(p<0.0001byKruskall-Wallistest). However,no difference was observed between the AS patients and controls.RA patients showed a significantly higherserum levels of Dkk1 than AS patients and controls (p<0.0001),but serum Dkk1 levelsin AS patientsweresignificantly lowerthan thosein controls (p<0.001).Thus,Dkk1/Rspo1ratiosweresignificantly elevated in RA patients (p<0.0001) and tended to decrease in AS patients when compared to the ratios in healthy controls. In RA patients, Dkk1/Rspo1 ratios better reflected the disease activity and joint damage status including ESR (p=0.030), CRP (p=0.004), DAS28 (p=0.049),erosive disease (p=0.026),and Sharp score (p=0.022)than each level.Moreover,after anti-TNF-α treatment,the ratios were significantly decreased (p=0.001 by Wilcoxon signed rank test)even though the changes of both Rspo1 and Dkk1 did not reach significance. However, Dkk1/Rspo1 ratio in AS patients was unchanged by anti-TNF-α treatment while Rspo1 levels were significantly increased (p=0.022).The Rspo1 and Dkk1 levels in SF from RA patients were significantly lower than those in serum samples(bothp<0.0001)andwerecomparabletothoseinSF from OA patients.
Conclusions:Thisstudy firstdemonstratedthatmeasurablelevelsof Rspo1 could be detected in human blood and thatDkk1/Rspo1 ratio,