key: cord-1025951-sw4crbdb authors: Sanchez-Rivera, L.; Iglesias, M. J.; Ibrahim-Kosta, M.; Kral- Pointner, J. B.; Havervall, S.; Goumidi, L.; Farm, M.; Munsch, G.; Germain, M.; Smith, P.; Hong, M.-G.; Suchon, P.; Naudin, C.; Boland, A.; Smadja, D. M.; Holmstrom, M.; Magnusson, M.; Silveira, A.; Uhlen, M.; Renne, T.; Martinez-Perez, A.; Emmerich, J.; Deleuze, J.-F.; Antovic, J.; Assinger, A.; Soria Fernandez, J. M.; Thalin, C.; Schwenk, J. M.; Souto Andres, J. C.; Morange, P.-E.; Butler, L. M.; Tregouet, D. A.; Odeberg, J. title: Elevated plasma Complement Factor H Regulating Protein 5 is associated with venous thromboembolism and COVID-19 severity date: 2022-04-21 journal: nan DOI: 10.1101/2022.04.20.22274046 sha: c1dc597191b33c33b58d3c4224c0dfe9a0103bdb doc_id: 1025951 cord_uid: sw4crbdb Venous thromboembolism (VTE), comprising both deep vein thrombosis (DVT) and pulmonary embolism (PE) is a common, multi-causal disease with potentially serious short- and long-term complications. In clinical practice, there is a need for improved plasma biomarker-based tools for VTE diagnosis and risk prediction. We used multiplex proteomics profiling to screen plasma from patients with suspected acute VTE, and a case-control study of patients followed up after ending anticoagulant treatment for a first VTE. With replication in 5 independent studies, together totalling 1137 patients and 1272 controls, we identify Complement Factor H Related Protein (CFHR5), a regulator of the alternative pathway of complement activation, as a novel VTE associated plasma biomarker. Using GWAS analysis of 2967 individuals we identified a genome-wide significant pQTL signal on chr1q31.3 associated with CFHR5 levels. We showed that higher CFHR5 levels are associated with increased thrombin generation in patient plasma and that recombinant CFHR5 enhances platelet activation in vitro. Thrombotic complications are a frequent feature of COVID-19; in hospitalised patients we found CFHR5 levels at baseline were associated with short-time prognosis of disease severity, defined as maximum level of respiratory support needed during hospital stay. Our results indicate a clinically important role for regulation of the alternative pathway of complement activation in the pathogenesis of VTE and pulmonary complications in acute COVID-19. Thus, CFHR5 is a potential diagnostic and/or risk predictive plasma biomarker reflecting underlying pathology in VTE and acute COVID-19. reflects the interplay between transient and sustained risk factors in disease development, 129 including acquired risk factors, genetics, and environmental exposures [24, 25] . VTE is a disease of the intravascular compartment and thus analysis of the blood proteome has 131 the potential to capture the resulting effects of combined genetic, epi-genetic, and environmental 132 contributors to risk variation. Large scale plasma proteomics screening allows for discovery of 133 novel protein biomarkers with potential clinical utility for diagnosis and/or prediction. Such studies 134 could also reveal biological pathways involved in VTE pathogenesis, for further functional studies 135 to identify targets for tailored treatment. So far, a handful of plasma proteomics studies of VTE 136 have been reported, presenting novel candidates associated with increased risk of VTE [26] [27] [28] [29] [30] , 137 including our own; the first affinity proteomics case-control study of patients followed up post 138 treatment after a first VTE, reporting PDGFB as a novel VTE associated biomarker [27] . High 139 throughput affinity proteomics studies have increased the capacity for discovery screening and 140 identification of novel associations, as now it is possible to screen for hundreds or thousands of 141 proteins in small plasma volumes [31] . Orthogonal verification is essential to confirm results 142 generated using these technologies but is frequently absent from such studies. Here, we aimed to identify novel biomarkers associated with acute VTE with a potential link to 144 underlying VTE pathogenesis. We identified complement factor H-related protein 5 (CFHR5), a 145 regulator of the alternative pathway of complement activation, as a novel VTE associated plasma 146 biomarker. Furthermore, we found that CFHR5 plasma levels are associated with short term 147 prognosis in acute COVID-19, a disease where thrombosis formation is central to pathology. Our 148 study suggests that CFHR5 could be involved in the underlying pathogenesis of VTE and acute 149 COVID-19, and that it is a potential clinical biomarker for thrombotic disease diagnosis and/or risk (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 Affinity plasma proteomics identifies candidate proteins associated with acute VTE To identify plasma biomarkers for VTE we analysed samples collected as part of our Venous 156 thromboEmbolism BIOmarker Study (VEBIOS) [27] . VEBIOS comprises two study arms; a 157 prospective cohort of patients sampled at the Emergency Room (ER), Karolinska University (Table 1) . Target candidates for measurement were selected as previously described [27], based on: (i) indications from the literature, in house data or public repositories of a probable 169 or plausible link to arterial or venous thrombosis (e.g., prior evidence of association with 170 thrombosis or intermediate traits, or known involvement in biological pathways of relevance), including 124 that we predicted to have endothelial enriched expression [32] , and (iii) the 172 availability of target specific antibodies in the Human Protein Atlas (HPA). A total of 758 HPA 173 antibodies, targeting 408 candidate proteins, were selected for incorporation into a single-binder 174 suspension bead array (Figure S1 A and Table S1, Tab_1), which was used to analyse plasma 175 generated from the blood samples collected into citrate anticoagulant. The signal generated by 176 antibody HPA059937, raised against the protein target sulfatase 1 (SULF1), was most strongly 177 associated with VTE (p<8.34E-06) ( Figure 1B , green point), with higher relative plasma levels in 178 cases vs. controls ( Figure 1D .i). Signals generated by a further seven antibodies were also associated with VTE (p<0.01) (Table S1, Tab_1). Protein signatures in plasma can be differently 180 affected by the sample matrix; anticoagulants can inhibit specific proteases, influence soluble 181 protein interactions, and modify analyte stability. Thus, the anticoagulant type used has potential 182 consequences for biomarker identification [33] . We therefore replicated the VEBIOS ER discovery 183 screen in EDTA samples drawn in parallel from the same patients ( Figure 1C ). Of the eight 184 antibodies that produced signals associated (p<0.01) with VTE in the citrate samples (Table S1, 185 Tab_1), four were replicated in the EDTA samples ( Figure 1B (Table S1 , Tab_1). In both 189 anticoagulants, all four target candidates were elevated in cases vs. controls (Figure 1D and E, . In all subsequent experiments, citrated blood was used in the analysis, on the basis that it is 191 the more commonly used sample anticoagulant in the clinical setting for coagulation analyses. Previously, in the VEBIOS Coagulation study (n=177) [27] , we identified 29 protein candidates in 193 plasma that were associated with prior VTE. This study was composed of patients sampled 1-6 194 months after discontinuation of anticoagulation treatment (duration 6-12 months) following a first 195 time VTE, or matched controls. Of the four antibodies that generated signals associated with 196 acute VTE in citrate and EDTA plasma in VEBIOS ER ( Figure 1B and C, marked with coloured 197 circles), only HPA059937 (predicted target SULF1), produced a signal associated with prior VTE 198 in the VEBIOS Coagulation study [27] . As our aim was to identify biomarkers associated with 199 acute VTE that were potentially linked to the underlying disease pathogenesis, we prioritised this 200 candidate on the basis that higher plasma concentrations observed in individuals with a 201 documented increased risk of VTE (e.g., previous VTE) in VEBIOS Coagulation study indicated 202 that it could also represent a constitutive and/or persistent risk factor. Complement Factor H Related protein 5 (CFRH5) is associated with VTE 204 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 The antibodies used in the single-binder suspension bead arrays passed quality control for 205 antigen binding specificity (see www.proteinatlas.org/), but selective binding to the target protein 206 in context of the complex matrix of plasma requires verification, as antibody specificity and 207 reliability can be a problematic issue [34] [35] [36] (Figure S1 C-F). To verify which protein(s) were 208 captured by HPA059937 (predicted target SULF1) we performed immunocapture-mass 209 spectrometry (IC-MS). Two proteins were bound to HPA059937 with a z-score>3 in triplicate p<0.00001]). We made five dual binder assays that targeted SULF1, but none gave a quantitative 226 signal in a plasma dilution series, or buffer containing a dilution series of recombinant SULF1 227 protein. These data are consistent with CFHR5, as opposed to SULF1, being the target protein 228 associated with VTE in the VEBIOS ER discovery screen. CFHR5 is associated with VTE independent of D-dimer or CRP 230 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 The CFHR5 dual binder assay using capture antibody HPA072446 ( Figure 1G .iii), together with 231 recombinant CFHR5 protein standard, was used for absolute quantification of CFHR5 in VEBIOS 232 ER and an extended sample set of the VEBIOS Coagulation study (n=284) ( We next investigated if CFHR5 levels were associated with VTE associated risk factors, such as 246 age, body mass index [BMI] ) and routine clinical laboratory tests for blood markers associated 247 with thrombosis risk (e.g., D-dimer, c-reactive protein [CRP], thrombocyte count). In VEBIOS ER, 248 CRP levels correlated with plasma CFHR5 concentration, in cases (⍴=0.57, p<0.001) and controls 249 (⍴=0.52, p<0.001) ( Figure 1H .ii), but there was no strong correlation between CFHR5 and the 250 other parameters measured, in cases or controls, including D-dimer (Figure 1H.iii and Table S1 , 251 Tab_3, Table A ). In the VEBIOS Coagulation study, CFHR5 levels in cases correlated with both 252 CRP (⍴=0.49, p<0.001) and D-dimer (⍴=0.42, p<0.0001) ( Figure 1J .ii and iii, right panel), but in 253 controls these correlations were weak (CRP ⍴=0.24, p=0.005) or absent, respectively ( Figure 254 1J.ii, and iii, left panel and Table S1 , Tab_3, Table B ). The association between CFHR5 and VTE 255 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 remained significant in both VEBIOS ER (p=0.0029) and VEBIOS Coagulation (p=0.0015) when 256 adjusted for CRP. Therefore, we did not adjust for CRP in further analyses. To further understand the expression characteristics of CFHR5, and to identify possible co-259 expressed or co-regulated proteins we used a whole transcriptome analysis approach. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. CFHR5 was most strongly linked to complement 3 (C3), the central hub of the largest of the three 285 linked cluster groups identified ( Figure 2C , clusters represented by green, red and cyan) ( The CFHR5 association with VTE replicates in additional cohorts The identification of biomarkers associated with VTE diagnosis, or risk profiling, requires 306 replication in independent cohorts, from different settings with different demographic profiles, to 307 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. ( Figure 3E ) (for all cohort details see Table S2 , Tabs 2-4). The relative risk of VTE associated 317 with CFHR5 per 1 SD increase in CFHR5 concentration was significant in all 3 replication cohorts: (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Figure 4A . Of N= 7,135,343 SNPs tested in a total sample of 2967 individuals, one 347 genome-wide significant (p < 5E-08) signal was observed on chr1q31.3. The lead SNP at this 348 locus was rs10737681, mapping to CFHR1/CFHR4, and the G allele was associated with β= 349 +81.33 ± 8.67 (p = 6.49E-21) increase in CFHR5 levels. After conditioning on the rs10737681, a 350 borderline significant association (p = 9.83E-08) was observed at the rs10494747 where the A-351 allele tended to associate with an increase of β= +47.60 ± 8.92 in CFHR5 levels. Rs10494747 is 352 distant of ~343 kb from CFHR2 rs10737681, maps to ZBTB41 ( Figure 4B ) and was in moderate 353 linkage disequilibrium (r 2~0 .10, D' = -0.63) with the lead rs10494747 variant. Together, these two 354 variants only explained 2.2% of the inter-individual variability in CFHR5 plasma levels. Neither 355 SNP was associated with VTE risk in the FARIVE and RETROVE case-control samples (Table 356 S1, Tab_7). 358 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. iii), although the effect appeared to be more strongly linked to stimulus concentration, than that 372 observed for ADP. Washed platelet response to ADP, convulxin or TRAP6 was not modified by 373 preincubation with CFHR5 ( Figure S2 ) (ANOVA all p>0.05), indicating that additional components 374 in plasma are required for the observed effects, and that they are not a direct effect of CFHR5 on 375 platelets. These data are consistent with the concept that CFHR5 is involved in the underlying 376 pathology of VTE. As thrombin generation has been associated with increased risk of VTE [55], we tested the 379 association between CFHR5 plasma concentration and thrombin generation as measured by 380 thrombinoscope in MARTHA (n=774 VTE cases, see Table S2 , Tab_5 for details) with replication 381 in RETROVE (308 cases/360 controls, see Table S2_Tab_4 for details). In both MARTHA and 382 RETROVE cases, we find significant association between CFHR5 and lag time (rho=0.181, p= 383 p<0.0001 and rho=0.176, p<0.0001, respectively), Endogenous Thrombin Potential (ETP) (rho=0.105, p= 0.0036, and rho=0.130, p<0.0001, respectively), peak (rho=0.117, p= 0.0012, and 385 rho=0.132, p<0.0001), and ttPeak (rho=0.116, p=0.0012, and rho=0.086, p=0.0274 (see Table 386 S1, Tab_8). All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101/2022.04.20.22274046 doi: medRxiv preprint Here, we aimed to identify biomarkers associated with acute VTE that are linked to disease 410 pathogenesis. Using a nested case-control study, derived from a cohort of patients presenting to 411 the ER with suspected acute VTE, and from a case-control study with patients that had suffered 412 a previous first VTE, we identify CFHR5, a regulator of the alternative complement activation 413 pathway, as such a biomarker. The association of CFHR5 with current or previous VTE was 414 replicated in three additional cohorts or case-control studies, and we also found a trend for 415 association with risk for recurrence of unprovoked VTE. We identify a genome-wide significant 416 signal associated with CFHR5 levels located in the CFHR1-5 gene cluster loci. We further provide 417 evidence of a direct role of CFHR5 in the induction of a pro-thrombotic phenotype, through its 418 effect on platelet activation. Finally, we show that increased CFHR5 was associated with degree 419 of respiratory insufficiency and prognosis in acute COVID-19. Our findings indicate CFHR5 has 420 potential application as a clinical biomarker for VTE diagnosis and risk prediction, providing further 421 support to the idea that complement regulation is a key element of VTE pathogenesis. Currently, D-dimer is the only plasma biomarker used in VTE diagnostic work-up, but its clinical 423 utility is limited to ruling out VTE in low-risk patients. Several studies have attempted to identify 424 novel biomarkers with potential clinically usefulness for the confirmation of VTE diagnosis, and 425 although a number have been identified [11] , none have yet been implemented in clinical practice. For many, like D-dimer, elevated levels are a consequence of thrombosis formation, e.g. biomarkers of fibrinolysis, clot re-modelling or resolution (e.g. MMPs), inflammation secondary to 428 local vascular and tissue injury (e.g., CRP, IL-6, IL-10, fibrinogen), or of endothelial and/or platelet 429 activation (e.g. vWF, P-selectin) [11, 60, 61] . We find no correlation or association between D-dimer and CFHR5 in the acute VTE setting, 431 supportive that increased concentration at diagnosis is not secondary to thrombus formation. In 432 contrast, we found a strong correlation between D-dimer and CFHR5 levels in patients followed 433 up after ending treatment for a first VTE, but not in controls. D-dimer has been associated with 434 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10. 1101 /2022 The CFHR5 locus maps to chromosome 1q31.3 at one end a gene cluster that spans 461 approximately 350 kb and that includes (in order from CFHR5) the CFHR2, CFHR4, CFHR1, 462 CFHR3, and CFH loci ( Figure 4B ). The rs10737681 we identify with genome wide association 463 with CFHR5 levels maps between the CFHR4 and CFHR1 genes ( Figure 4B) Our study indicates that CFHR5 has a possible role in the potentiation of platelet activation. Using (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Our study has various strength and limitations; VEBIOS ER, the discovery cohort, that was 506 derived from a single centre, where blood sampling for plasma biobanking was performed in 507 parallel to that for routine tests after initial evaluation (before diagnostic imaging or anticoagulant 508 treatment), thus avoiding bias in inclusion or biobanking. Samples were handled according to 509 standard clinical chemistry lab routine, thus variations in needle-to-spin-to-freeze time were 510 equivalent between case and control samples. As biobanking was based on the routine sample 511 flow, this increases the feasibility that identified biomarker candidates are suitable for clinical 512 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101/2022.04.20.22274046 doi: medRxiv preprint translation into a routine setting. Importantly, we demonstrate an association of CFHR5 with VTE 513 in several independent studies, that include patients in the acute setting, at follow up, and prior to 514 recurrence. One limitation of our study is that we have not analysed a cohort of individuals that 515 were sampled prior to VTE event. From a technological perspective, our study demonstrates the 516 need for orthogonal verification of any potential biomarker identified using antibody-based 517 proteomics screening [34, 35] . The same caution should be extended to findings generated using 518 other high throughput affinity proteomics technologies vulnerable to non-specific protein binding, 534 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Hospital in Huddinge, Stockholm, as previously described [47] . The patients were out-patients 577 with low-to-high probability of acute PE or DVT in a lower limb. The study was approved by the Table S2 , Tab_2. The FARIVE study is a French multicentre case-control study carried out between 2003-2009, as 585 previously described [77] . The study consists of patients with first confirmed VTE (DVT to the 586 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Table S2 , Tab_4. The Marseille Thrombosis Association study (MARTHA) is a population based single centre study, (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 which was used to analyse the association between CFHR5 and blood coagulability. Clinical 612 characteristics are described in Table S2 , Tab_5. The COVID-19 biomarker and Immunity (COMMUNITY) study is a single centre study of 112 Table S2 , Tab_6. Plasma proteomic profiles in VEBIOS ER were generated using multiplexed suspension bead 633 arrays (SBA) with 758 antibodies selected from the Human Protein Atlas (HPA) project, targeting 634 408 proteins (Table S1 ), using identical design, procedures and methods as previously described 635 [27] . Briefly, paired samples were randomly distributed within the same 96-well area. Two 636 suspension bead arrays composed of 380 antibodies and 4 controls were used to sequentially 637 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 anti-human CFHR5 (R&D systems, MAB3845) antibody were labelled with biotin and used as Analysis of CFHR5 in COVID-19 patients in the COMMUNITY study 689 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 The HPA antibody HPA073894, targeting the CFHR5 protein, was included in a screening panel 690 using multiplexed suspension bead arrays (SBA) following a similar protein profiling protocol as 691 in the targeted plasma proteomics screening (see above). Association of CFHR5 plasma levels 692 with Respiratory support Index (RI) at baseline was tested using linear regression models. RI (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 platelets were incubated with recombinant CFHR5 in PBS (6 µg/ml, 3845-F5, R&D systems) or 742 PBS alone for 10 minutes before treatment with varying concentrations of ADP (1-5 µM), TRAP- (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 lower than 20, 1,774 variants with a call rate lower than 95%. Finally, 535,105 markers passed 768 the control quality and were used for the imputation. The Imputation was performed with Minimac4 769 using the 1000 Genomes phase 3 version 5 reference panel [87] . All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 The study was supported by grants from Stockholm City Council (SLL) to JO (2017-0842/0587, All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101 /2022 (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101/2022.04.20.22274046 doi: medRxiv preprint Figure 1 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101/2022.04.20.22274046 doi: medRxiv preprint Figure 1 : Plasma proteomics profiling identifies CFHR5 as associated with VTE (A) Overview of the VEBIOS ER discovery cohort. 758 HPA antibodies, targeting 408 candidate proteins, were used to analyse plasma samples using affinity proteomics. Log fold changes in antibody MFI (mean fluorescent intensity) signal were calculated between VTE cases and controls in (B) citrate or (C) EDTA anticoagulated plasma; coloured circles indicate antibodies that generated signals significantly associated with VTE in both. MFI signals generated by these antibodies for controls and cases in (D) citrate plasma and (E) EDTA plasma. (F) Immunocapture-mass spectrometry identification of protein targets of HPA059937. (G) Dual binder assays were developed using an anti-CFHR5 detection antibody, combined with (i) HPA059937 (raised against SULF1) (ii), anti-CFHR5 HPA073894 or (iii) anti-CFHR5 HPA072446 as capture antibodies. CFHR5 levels in the citrated plasma samples were re-analysed, using the respective dual binder assays, to determine (vii-ix) levels (MFI) in controls vs. cases and (vii-ix) the correlation between the signal and those generated by the original single binder assay using HPA059937. Dual binder assay using capture antibody HPA072446 with a recombinant protein standard was used for absolute quantification of CFHR5 in samples from: (H) VEBIOS ER and (J) VEBIOS Coagulation. CFHR5 concentration was (i) measured in controls and cases, with associated OR (odds ratio per 1 standard deviation increase) or (ii) used to determine the correlation with C-reactive protein (CRP), or (iii) d-dimer concentration. ***p<0.0001, ***p<0.001, **p<0.01. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. concentration was measured in plasma from (D) VEBIOS ER or (E) VEBIOS coagulation to determine: (i) differences between controls and cases, or (ii) correlation with CFHR5 in controls (left) or cases (right). **p<0.01. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The French FARIVE study recruited patients with confirmed acute VTE, with controls recruited from hospital patients treated for non-VTE causes. Samples were drawn within one week from diagnosis, during initiation of treatment. (D) The Swedish VEBIOS Coagulation or (E), Spanish RETROVE study recruited cases from patients who had a prior first time VTE, sampled post-treatment (6-12 months anticoagulants), with healthy controls recruited from the general population. CFHR5 concentration was measured in the respective samples using a dual binder assay **p<0.01, ***p<0.001, ****p<0.001 OR (1SD) = Odds ratio for 1 standard deviation elevation. CI= confidence interval. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 The lead SNP at this locus, rs10737681, maps between the CFHR1 and CFHR4 gene loci in the gene cluster of CFHR1-5. A borderline significant association (p = 9.83E-08) with CFHR5 levels was observed at the rs10494747, mapping to the ZBTB41 gene. Indicated in green is the rs143410348 recently identified with genome wide significance as associated with VTE risk [68] . All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101/2022.04.20.22274046 doi: medRxiv preprint All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101/2022.04.20.22274046 doi: medRxiv preprint COVID-19 were included prospectively and followed longitudinally for disease severity (ii). Plasma was sampled at inclusion and, on average, every second day during hospital stay. (iii) In total, 359 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. samples to analyse the relationship with respiratory support index (RI) recorded at the time of sampling or (ii) in the 112 baseline samples to analyse the relationship with the maximum RI for that patient recorded at any subsequent time point (ii). *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; DVT, deep vein thrombosis; PE, pulmonary embolism; n, numbers; SD, standard deviation; †, within one month from diagnosis or index date (immobilization with trauma, surgery, cast and/or orthosis and bedrest more than 3 days of sickness); ‡, within the last year; *, on-going treatment; IQR, interquartile range; CFHR5, Complement Factor H-related protein 5; C3, Complement 3; CRP, Creactive protein; LPK, leucocytes; Hb, haemoglobin; TPK, thrombocytes. Missing values across all variables are <10%. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted April 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 Incidence of venous thromboembolism in 804 northern Sweden (VEINS): a population-based study Epidemiology of venous thromboembolism 30-year mortality after venous thromboembolism: a population-808 based cohort study. Circulation Epidemiology of first and recurrent venous thromboembolism: A 810 population-based cohort study in patients without active cancer. Thrombosis and 811 Haemostasis Evaluation of D-dimer in the diagnosis of suspected deep-vein 813 thrombosis Excluding pulmonary embolism at the bedside without diagnostic 815 imaging: management of patients with suspected pulmonary embolism presenting to the 816 emergency department by using a simple clinical model and d-dimer CT pulmonary angiography: an over-utilized imaging modality in 819 hospitalized patients with suspected pulmonary embolism D-Dimer Use and Pulmonary Embolism Diagnosis in Emergency 825 Units: Why Is There Such a Difference in Pulmonary Embolism Prevalence between the 826 United States of America and Countries Outside USA? PLoS One Highly Elevated Quantitative D-Dimer Assay Values Increase the 829 Likelihood of Venous Thromboembolism. TH Open Diagnostic biomarkers in venous thromboembolic 831 disease Risk assessment of recurrence in patients with unprovoked deep vein 833 thrombosis or pulmonary embolism: the Vienna prediction model Identifying unprovoked thromboembolism patients at low risk for 842 recurrence who can discontinue anticoagulant therapy Unprovoked recurrent venous thrombosis: prediction by D-dimer and 845 clinical risk factors No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted The Human Plasma Proteome Draft of 2017: Building on the 893 Human Plasma PeptideAtlas from Mass Spectrometry and Complementary Assays. J 894 Proteome Res The PeptideAtlas project The Genotype-Tissue Expression (GTEx) project A single-cell type transcriptomics map of human tissues Gene ontology: tool for the unification of biology The Gene Ontology resource: enriching a GOld mine. Nucleic Acids 906 Res WGCNA: an R package for weighted correlation network 908 analysis Age-adjusted D-dimer cut-off leads to more efficient diagnosis of venous 916 thromboembolism in the emergency department: a comparison of four assays Association of influenza vaccination with reduced risk of venous 919 thromboembolism Genome wide association study for plasma levels of natural 923 anticoagulant inhibitors and protein C anticoagulant pathway: the MARTHA project Functional Relevance of the Anaphylatoxin Receptor C3aR for 926 Platelet Function and Arterial Thrombus Formation Marks an Intersection Point 927 Between Innate Immunity and Thrombosis. Circulation Distinct contributions of complement factors to platelet 930 activation and fibrin formation in venous thrombus development Human factor H-related protein 5 has cofactor activity, inhibits C3 933 convertase activity, binds heparin and C-reactive protein, and associates with 934 lipoprotein D-dimer, thrombin generation, and risk of a first venous thrombosis in 936 the elderly Complement associated microvascular injury and thrombosis in the 938 pathogenesis of severe COVID-19 infection: A report of five cases Prothrombotic changes in patients with COVID-19 are 947 associated with disease severity and mortality Inflammatory biomarkers in deep venous thrombosis organization, 950 resolution, and post-thrombotic syndrome The risk of a first and a recurrent venous thrombosis 955 associated with an elevated D-dimer level and an elevated thrombin potential: results of 956 the THE-VTE study Regulation of regulators: Role of the complement factor H-related 960 proteins Factor H and disease: a complement regulator affects vital body 962 functions Factor H-related protein-5: a novel component of human glomerular 964 immune deposits Factor H-related protein 5 interacts with pentraxin 3 and the 966 extracellular matrix and modulates complement activation Complement and coagulation: strangers or partners in crime? 969 Trends Immunol Time Course of a Second Outbreak of COVID-19 in Beijing, China Endothelial cell infection and endotheliitis in COVID-19 Increased complement activation is a distinctive feature of severe SARS-977 CoV-2 infection. bioRxiv Incidence of thrombotic complications in critically ill ICU patients with 979 COVID-19 Synergistic insights into human health from aptamer-and antibody-984 based proteomic profiling Soluble Urokinase Plasminogen Activator Receptor: Genetic 986 Variation and Cardiovascular Disease Risk in Black Adults Common susceptibility alleles are unlikely to contribute as 991 strongly as the FV and ABO loci to VTE risk: Results from aGWAS approach Platelet count and plateletcrit are associated with an 994 increased risk of venous thrombosis in females. Results from the RETROVE study A meta-analysis of genome-wide association studies identifies 997 ORM1 as a novel gene controlling thrombin generation potential Probabilistic quotient normalization as robust method to account for 1000 dilution of complex biological mixtures. Application in 1H NMR metabonomics Multidimensional Normalization to Minimize Plate Effects of 1003 Suspension Bead Array Data Human genomics. The Genotype-Tissue Expression (GTEx) pilot 1009 analysis: multitissue gene regulation in humans Genetic Determinants of Thrombin Generation and Their 1011 Relation to Venous Thrombosis: Results from the GAIT-2 Project Next-generation genotype imputation service and methods Genetics of venous thrombosis: insights from a new genome wide 1016 association study Combined analysis of three genome-wide association studies on vWF 1018 and FVIII plasma levels Meta-analysis of 65,734 individuals identifies TSPAN15 and 1020 SLC44A2 as two susceptibility loci for venous thromboembolism GWAMA: software for genome-wide association meta-1023 analysis (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.