The Functional Polymorphism Ala258Ser in the InnateReceptor Gene Ficolin-2 in the Donor Predicts

Improved Renal Transplant Outcome

Michael Eikmans,1,5 Ilse de Canck,2 Pieter van der Pol,3 Carla C. Baan,4 Geert W. Haasnoot,1

Marko J.K. Mallat,3 Manon Vergunst,1 Els de Meester,2 Joke I. Roodnat,4 Jacqueline D.H. Anholts,1

Martine van Thielen,2 Ilias I.N. Doxiadis,1 Johan W. de Fijter,3 Pieter J.E. van der Linden,1

Els van Beelen,1 Cees van Kooten,3 Judith A. Kal-van Gestel,4 Annemiek M.A. Peeters,4

Willem Weimar,4 Dave L. Roelen,1 Rudi Rossau,2 and Frans H.J. Claas1

Background. Innate immunity plays a role in controlling adaptive immune responses.Methods. We investigated the clinical relevance of single nucleotide polymorphisms in 22 genes encoding innate,secreted, and signaling pattern recognition receptors in a total of 520 donor-recipient pairs of postmortem, humanleukocyte antigenYDR-compatible kidney transplantations. Associations with rejection incidence were tested in ana priori randomized training set and validation set.Results. Polymorphisms in TLR-3 (rs3775296) in the recipients and in Ficolin-2 (rs7851696; Ala258Ser) and C1qR1(rs7492) in the donors showed the strongest association with severe rejection. In multivariate analysis, presence ofthe Ficolin-2 Ala258Ser variant in the donor predicted lower incidence of severe rejection (odds ratio=0.3; 95%confidence interval, 0.1Y0.9; P=0.024) and of graft loss (hazard ratio=0.5; 95% confidence interval, 0.2Y1.0; P=0.046)independently of clinical risk factors. Ficolin-2 messenger RNA expression was detected in pretransplantation bi-opsies from 69 donor grafts. Serum and tissue Ficolin-2 levels were unaffected by genotype. Ficolin-2 protein, whichbound to dying cells, was detected in donor kidneys in a passenger leukocyte-like pattern. Indeed, monocytes,monocyte-derived macrophages, and peripheral blood mononuclear cells expressed Ficolin-2. Donor grafts withthe Ficolin-2 Ala258Ser variant contained significantly elevated expression of interleukin 6, having ascribed cyto-protective effects. It has been described that Ala258Ser leads to increased binding capacity of Ficolin-2 to N-acetylglucosamine.

Parts of this work have been financially supported by the Dutch KidneyFoundation (grants C07-2238 and C03-6014).

The authors declare no conflicts of interest.1 Department of Immunohematology and Blood Transfusion, Leiden Uni-

versity Medical Center, Leiden, the Netherlands.2 R&D Discovery, Innogenetics, Gent, Belgium.3 Department of Nephrology, Leiden University Medical Center, Leiden, the

Netherlands.4 Department of Internal Medicine, Erasmus University Medical Center,

Rotterdam, the Netherlands.5 Address correspondence to: Michael Eikmans, Ph.D., Department of Im-

munohematology and Blood Transfusion, Leiden University MedicalCenter, Bldg. 1, E3-Q, Albinusdreef 2, 2333 ZA, Leiden, the Netherlands.

E-mail: m.eikmans@lumc.nlM.E. participated in analyzing and interpreting the data, making the research

design, and writing the article. I.d.C. participated in analyzing andinterpreting the single nucleotide polymorphism data, interpretingstatistical data, and making the research design. P.v.d.P. participated indeveloping cell death and Ficolin-2 binding assays and performing andinterpreting Ficolin-2 immunohistochemistry. C.C.B. participated inmanaging databases and generating patient data for the Rotterdam co-hort. G.W.H. participated in performing and interpreting statisticalanalyses. M.J.K.M. participated in managing and generating patient datafor the Leiden cohort. M.V. participated in performing and interpretingcell death and Ficolin-2 binding assays. E.d.M. participated in developingand optimizing single nucleotide polymorphism assays and making theresearch design. J.I.R. participated in generating patient data for theRotterdam cohort. J.D.H.A. participated in performing and interpreting

results from polymerase chain reaction assays, single nucleotide poly-morphism analyses, and messenger RNA expression measurements.M.v.T. participated in analyzing and interpreting the single nucleotidepolymorphism data. I.I.N.D. participated in managing patient databasesand interpreting the statistical data. J.W.d.F. participated in generatingpretransplantation biopsies, patient data, and material for the Leidencohort. P.J.E.v.d.L. participated in performing and interpreting poly-merase chain reaction assays and single nucleotide polymorphismanalyses. E.v.B. participated in performing and interpreting Ficolin-2assays in serum. C.v.K. participated in generating pretransplantationbiopsies and interpreting the data. J.A.K.-v.G. and A.M.A.P. participatedin managing the database for the Rotterdam cohort. W.W. participatedin generating patient data and material for the Rotterdam cohort. D.L.R.participated in generating the patient serum. R.R. participated indirecting development of single nucleotide polymorphism assays andmaking the research design. F.H.J.C. participated in making the researchdesign and interpreting the data. All authors participated in reviewingthe article and approving it for submission.

Supplemental digital content (SDC) is available for this article. Direct URLcitations appear in the printed text, and links to the digital files areprovided in the HTML text of this article on the journal’s Web site(www.transplantjournal.com).

Received 31 January 2012. Revision requested 28 February 2012.Accepted 25 April 2012.Copyright * 2012 by Lippincott Williams & WilkinsISSN: 0041-1337/12/9405-478DOI: 10.1097/TP.0b013e31825c5967

CLINICAL AND TRANSLATIONAL RESEARCH

478 www.transplantjournal.com Transplantation & Volume 94, Number 5, September 15, 2012

Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Conclusions. Presence of the Ficolin-2 Ala258Ser polymorphism in the donor independently predicts improved graftoutcome. Based on mechanistic data, we propose that this functional polymorphism leads to more efficient handlingof injured cells by phagocytozing cells, resulting in decreased intragraft exposure to danger signals and dampenedalloimmune responses.

Keywords: Kidney transplantation, Acute rejection, Graft outcome, Innate immunity, DNA polymorphism, Ficolin,Cell death.

(Transplantation 2012;94: 478Y485)

T he degree by which an immune response is induced insolid organ transplants varies between individuals. This

variation is determined partly by DNA sequence differencesin promoter and coding regions of immunologically rele-vant genes. To reduce the risk of rejection, the HLA genotypeof the donor and the recipient should be as similar as pos-sible. Because even in perfectly human leukocyte antigen(HLA)Ymatched donor recipient pairs rejection can occur,other markers including innate immunity genes are likely toplay a role as well.

The innate immune system acts as a first line of defenseagainst infections and is involved in clearance of potentiallydangerous substances arising from cellular debris and dam-aged cells within the host (1). Invading pathogens give riseto pathogen-associated molecular patterns that are detectedby pattern recognition receptors (PRRs). In a similar man-ner, danger signals or damage-associated molecular patterns,which are released by damaged or stressed cells, can be boundby PRRs (1). The PRRs are present in the cell membrane orin soluble form in the circulation. The group of cell-boundreceptors is formed by the Toll-like receptors (TLRs) (2).The group of soluble recognition molecules includes C1q,mannose-binding lectin (MBL), and the ficolins.

Mechanistic paradigms have been focused on the roleof adaptive immunity in graft rejection. More recently, therole of innate PRRs in rejection of solid organ transplantshas been emphasized (3Y6). In kidney transplantation,antigen-independent insult to the donor organ can inducePRR signaling and priming of the alloimmune response.Ischemia-reperfusion damage and oxidative stress lead tothe release of high-mobility group box (HMGB)Y1, heparansulfate, hyaluronan, fibrinogen, and heat-shock proteins fromdead cells, which act as endogenous ligands activating PRRs(7, 8). On ischemic injury, PRRs potentially modulate theextent of inflammation (9, 10). MBL, Ficolin-2, Ficolin-3,and C1q can engage apoptotic and necrotic cells (11Y14) and,in this way, participate in the removal of dying cells andmaintenance of tissue homeostasis.

In kidney transplantation, data on the clinical relevanceof single nucleotide polymorphisms (SNPs) in PRR genes arelimited. Donor kidneys with a TLR-4 loss-of-function allele,conferring diminished affinity for HMGB1, have lower ex-pression of proinflammatory cytokines and have a higherchance of immediate graft function (10). TLR-4 gene poly-morphisms were associated with reduced incidence of acuterejection in lung (15) and kidney transplants (16, 17). Su-perior graft survival in kidney and kidney-pancreas trans-plantation is associated with low MBL serum levels (18, 19)resulting from a SNP in exon 1 of the MBL2 gene (18).

In the current study, we investigated the clinical rele-vance of SNPs in PRRs within cohorts of donor-recipientcombinations of kidney transplantations.

RESULTS

Association of PRR Polymorphisms withRejection Incidence

Demographics of study cohorts and genotype distri-bution of SNPs have been summarized in Table 1 and Sup-plemental Table 1 (see SDC, http://links.lww.com/TP/A681),respectively. In both the test and validation cohort, a SNPin the 5¶-untranslated region of TLR-3 (rs3775296) in therecipients and an amino acid substituting SNP in exon 8 ofthe Ficolin-2 gene (rs7851696; Ala258Ser; GT variant) in thedonors showed association with incidence of antithymocyteglobulin (ATG)Yrequiring rejection (Table 2) (see Table 2,SDC, http://links.lww.com/TP/A681). SNPs in the promoterregion and exon 2 of TLR-9 showed association with out-come (Table 2), but genotype distribution differed betweenthe training and validation sets.

In addition to the replication-based analysis, a joint testwas performed (Table 2). Strongest associations were foundfor recipient TLR-3 (P=0.0015), donor Ficolin-2 (P=0.0029),and donor C1qR1 (P=0.0007) with ATG-requiring rejec-tion. Handling a false discovery rate at 5%, only the latterremained significant.

Donor Ficolin-2 Genotype PredictsGraft Outcome

In multivariate logistic regression, the donor Ficolin-2Ala258Ser variant predicted lower risk of severe rejectionindependently of clinical variables (odds ratio=0.3; 95%confidence interval, 0.1Y0.9; P=0.024) (Table 3). In multi-variate Cox regression analysis (Table 3), the Ala258Servariant predicted lower rate of graft loss (hazard ratio=0.5;95% confidence interval, 0.2Y1.0; P=0.046). Patients who hadreceived a graft from a donor with the Ficolin-2 Ala258Ser(GT) variant showed significantly better graft survival (79.8%)than recipients whose donor had the Ficolin-2 wild-type (GG)variant (58.1%, P=0.016, Fig. 1). Within the first 3 months,almost 10% of the grafts in the group with Ficolin-2 wild-typedonors were lost, and donor genotypes differed 12.2% in graftsurvival (P=0.016). Permanent nonfunction and thrombosis/infarction accounted for most (72%) of the losses. With thegraft survival set to 100% at the 1-year time point, survivalplots for groups with different Ficolin-2 genotypes still di-verged, accounting for another 10% survival difference at8 years after transplantation. The TLR-3, TLR-9, and C1qR1genotype did not impact graft survival (data not shown).

Further Investigation of Ficolin-2

Ficolin-2 Serum Levels are Not Affected by Genotype

Ficolin-2 levels in serum of patients with the wild-typevariant (5.3 T 4.2Kg/mL) were not significantly different fromthose in patients with the Ala258Ser variant (6.0 T 4.4Kg/mL).Likewise, investigation of Ficolin-2 levels in serum from

* 2012 Lippincott Williams & Wilkins Eikmans et al. 479

Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

donors did not show differences between genotypes (wildtype: 8.8 T 3.9 Kg/mL; Ala258Ser: 8.0 T 4.4 Kg/mL).

Donor Grafts Express Ficolin-2

Messenger RNA expression for Ficolin-2 was detectedin 69 pretransplantation biopsies studied. The average Ct

value was 28.3 T 1.6, indicating detectable transcriptionwithin the donor kidneys. Average Ct values for referencegenes A-actin and GAPDH were 18.0 T 1.6 and 18.8 T 1.8,respectively. Ficolin-2 messenger RNA (mRNA) levels in thepretransplantation biopsies were not affected by genotype(wild type: 1 T 0.89; Ala258Ser: 0.69 T 0.38; P=not signifi-cant). Ficolin-2 protein expression was detected in donorkidneys between the tubules (Fig. 2A). The expression

pattern seemingly resembles that of passenger leukocytes.Significant amounts of Ficolin-2 mRNA were detected inperipheral blood mononuclear cells (PBMCs), CD14+ mono-cytes, and differentiated macrophages (Fig. 2B).

Ficolin-2 Binds to Dying Cells

Ficolin-2, possibly derived from local macrophages inthe graft, may be involved in recognition of apoptotic cells.As an in vitro model, dying Jurkat T cells were incubatedwith recombinant Ficolin-2 protein, which bound to lateapoptotic cells (Annexin-A5pos and 7-AADpos) and, to alesser degree, to early apoptotic cells (Annexin-A5pos and7-AADneg) (Fig. 2C).

Intragraft Cytokine Expression in Relation toFicolin-2 Genotype

Interleukin (IL)-6 has a cytoprotective effect inischemia-reperfusion injury (20, 21). Expression of IL-6 wasabundant in pretransplantation biopsies and was signifi-cantly (P=0.023) higher in donor grafts with the Ficolin-2

TABLE 1. Demographics of study cohortsa

VariableTraining Set

(n=189)Validation Set

(n=302) P

Period of transplantation NS

e1996 46.9 49.6

91996 53.1 50.4

Transplantation number NS

1 82.1 83.0

91 17.9 17.0

Transplant center NS

Leiden 53.1 52.8

Rotterdam 46.9 47.2

Current antibodies NS

e5% 74.5 75.2

95% 25.5 24.8

Highest antibodies NS

e5% 37.2 36.2

95% 62.8 63.8

Cold ischemia time NS

e24 hr 59.5 60.3

924 hr 40.5 39.7

HLA-A, B mismatching NS

No 35.7 33.0

Yes 64.3 67.0

Recipient gender NS

Male 59.2 54.3

Female 40.8 45.7

Donor gender NS

Male 63.8 65.6

Female 36.2 34.4

Recipient age NS

e50 yr 59.2 54.3

950 yr 40.8 45.7

Donor age NS

e50 yr 63.8 65.6

950 yr 36.2 34.4

Rejection rate 29.5 31.7 NS

ATG-requiring rejection 18.5 19.9 NS

a Table shows percentages within cohorts.ATG, antithymocyte globulin; HLA, human leukocyte antigen; NS, not

significant.

TABLE 2. Associations between SNPs and kidney graftrejection incidencea

Gene Rs No.Training

SetValidation

SetJoint

Set

Rejection (Yes/No)

Recipient

TLR-3 (5’-UTR) rs3775296 0.025 0.075 0.0058

TLR-8 (exon 3) rs2159377 0.037 0.424 0.155

TLR-9 (promoter) rs187084 0.028 0.259 0.043

CD14 (exon 2) rs2563298 0.037 0.942 0.200

Masp-2 (intron 8) rs6695096 0.034 0.147 0.823

Donor

TLR-1 (exon 4) rs4833095 0.033 0.614 0.156

TLR-5 (exon 6) rs2072493 0.039 0.306 0.040

TLR-9 (promoter) rs187084 0.046 0.030 0.012

TLR-9 (exon 2) rs352140 0.024 0.023 0.0051

LBP (exon 10) rs1780627 0.024 0.326 0.025

C1qR1 (3’-UTR) rs7492 0.006 0.234 0.0079

ATG-Requiring Rejection (yes/no)

Recipient

TLR-3 (5’-UTR) rs3775296 0.037 0.018 0.0015

CD14 (exon 2) rs2563298 0.020 0.821 0.149

Donor

LBP (promoter) rs2232578 0.021 0.818 0.122

LBP (exon 10) rs1780627 0.013 0.228 0.035

LBP (exon 13) rs2232618 0.007 0.950 0.113

Ficolin-1 (exon 9) rs1071583 0.024 0.998 0.273

Ficolin-2 (exon 8) rs7851696 0.042 0.028 0.0029

C1qR1 (3’-UTR) rs7492 0.001 0.083 0.0007

a The table shows P values. Only SNPs with a significant association in thetraining set are shown.

LBP, lipopolysaccharide-binding protein; Masp, mannose-associatedserine protease; prom, promoter; TLR, Toll-like receptor; UTR, untrans-lated region.

480 www.transplantjournal.com Transplantation & Volume 94, Number 5, September 15, 2012

Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Ala258Ser variant than in those with the wild-type variant(Fig. 2D).

DISCUSSIONWe investigated the clinical relevance of polymorph-

isms in genes encoding innate PRRs in kidney transplanta-

tion. The donor Ficolin-2 genotype predicted both incidenceof severe rejection and graft survival independently fromclinical variables.

Because replication of findings from molecular studiesin kidney transplantation is needed (22), we pursued a two-stage approach by investigating a randomized training andvalidation set. Second, because joint analysis can be statis-tically more powerful than replication-based analysis (23),cohorts were also combined, and association were tested witha false discovery rate at 5%. After correction for multipletesting, P values for associations of the Ficolin-2 SNP withsevere rejection incidence exceeded 0.05. Despite this, donorFicolin-2 genotype independently predicted graft outcome.We think that the effect of donor Ficolin-2 genotype is dual:a relatively early effect within the first 3 months and aneffect on graft survival in the following years, which is prob-ably influenced by the different incidence of severe (ATG-requiring) rejection between genotype variants. Indeed, graftsurvival curves for recipients having ATG-requiring rejec-tion and recipients with no need for ATG start to dissociateat 1.5 year after transplantation (data not shown). Ficolin-2polymorphisms have not been investigated in kidney trans-plants before. SNP rs3775296 in recipient TLR-3 was sig-nificantly associated with incidence of severe rejection. Suchassociation was not found previously (24), possibly becauseonly living-donor kidney transplantations had been studied.A significant association between recipient TLR-3 genotypeand acute rejection incidence was found in a recent study(25), where the majority had been transplanted with a graftfrom a deceased donor.

Differences in the impact on graft outcome betweenFicolin-2 genotypes might be explained by a difference insusceptibility to infection, whereby liver-derived Ficolin-2in the recipient is expected to be involved in combating

TABLE 3. The Ficolin-2 Ala258Ser (GT) variant in the donor independently predicts improved kidney graft outcome

Univariate Multivariate

OR (95% CI) P OR (95% CI) P

Logistic Regression (Occurrence of Severe Rejection)Donor Ficolin-2 Ala258Ser (GT) variant 0.4 (0.1Y0.9) 0.034 0.3 (0.1Y0.9) 0.024

Donor C1qR1 CT variant 1.9 (0.9Y3.9) 0.087 V

Recipient TLR-3 CA variant 1.7 (0.9Y3.2) 0.078 V

Recipient age (older than 50 yr) 0.5 (0.3Y0.8) 0.010 0.4 (0.2Y0.8) 0.013

Period of transplantation (later than 1996) 0.4 (0.2Y0.7) 0.001 0.3 (0.1Y0.6) G0.001

Cox Regression (Graft Loss) HR (95% CI) P HR (95% CI) P

Donor Ficolin-2 Ala258Ser (GT) variant 0.4 (0.2Y0.9) 0.020 0.5 (0.2Y1.0) 0.046

Donor C1qR1 CT variant 0.9 (0.5Y1.7) 0.91 V

Recipient TLR-3 CA variant 0.9 (0.5Y1.4) 0.59 V

Donor age (older than 50 yr) 2.3 (1.5Y3.5) G0.001 2.6 (1.6Y4.1) G0.001

Transplant center 1.9 (1.2Y2.9) 0.004 1.8 (1.1Y2.9) 0.012

HLA class I mismatching 1.7 (1.0Y2.7) 0.036 1.8 (1.1Y3.1) 0.020

Number of transplants (more than 1) 1.6 (1.0Y2.7) 0.061 2.0 (1.2Y3.4) 0.011

Period of transplantation (later than 1996) 0.7 (0.4Y1.1) 0.077 0.6 (0.3Y1.0) 0.043

CI, confidence interval; HLA, human leukocyte antigen; HR, hazard ratio; OR, odds ratio; TLR, Toll-like receptor.

FIGURE 1. Association between the donor Ficolin-2Ala258Ser (GT) variant andhigher rateof kidneygraft survival.The genotype for Ficolin-2 (wild-type/GG variant: dotted line;Ala258Ser/GT variant: straight line) in the donorwas tested forassociation with graft loss (censored for death with functioninggraft) using Kaplan-Meier curves and log-rank testing.

* 2012 Lippincott Williams & Wilkins Eikmans et al. 481

Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

pathogens. This possibility seems less likely because it wasthe donor genotype that determined kidney graft outcome.Moreover, in liver transplantation, the Ala258Ser SNP doesnot increase risk of infection (26). We do not think thatantibody formation against the exon-8 region of Ficolin-2explains the difference in clinical outcome between geno-types because Ficolin-2 genotype donor-recipient incom-patibility did not result in worse graft survival compared withmatched combinations (data not shown).

We hypothesize that the SNP in Ficolin-2, expressedby phagocytozing cells in the engrafted kidney, enables moreefficient handling of injured cells, leading to decreased ex-posure to danger signals and dampened immune responses.A model is depicted in Supplemental Figure 1 (see SDC,http://links.lww.com/TP/A681), based on the concept ofclearance of dying cells and its immunologic consequences(27Y30). Apoptotic cells not timely removed by appropriate

opsonization and phagocytosis may show secondary ne-crosis. Cell membranes are disrupted, and harmful intra-cellular components are exposed on the surface, which mayinduce inflammation by activating the inflammasome (31)and inducing TLR and NF-JB signaling in macrophages.

The previous concept is supported by the fact that coldpreservation and transplantation of the donor kidney resultinto higher numbers of dead cells (32, 33). Increased numbersof apoptotic parenchymal cells in the donor kidney beforeengraftment are related to and predictive of decreased renalfunction early after transplantation (34, 35). Dampening ofdanger signal HMGB1, exposed by tubular epithelial cellsduring ischemia (10), leads to reduction in tubular apoptosisand inflammation and to preservation of kidney function(36). Ficolin-2 facilitates phagocytosis by binding to dyingcells (this article and those of Kuraya et al. (13) and Jensenet al. (14)). In the circulation, the binding of Ficolin-2 to

FIGURE 2. Ficolin-2 expression and binding studies. A, Protein staining of Ficolin-2 in a cadaveric donor kidney. Fourdifferent kidneys were analyzed. Ficolin-2 protein (red arrows) was detected between the tubules. GL indicates glomerulus.B, Ficolin-2 messenger RNA (mRNA) expression was investigated in pretransplantation biopsies (pre-Tx Bx), spleen tissue,peripheral blood mononuclear cells (PBMCs), cultured dendritic cells (DCs), CD14+ monocytes, differentiated (macro-phage-colony stimulating factor or granulocyte macrophageYcolony stimulating factor) macrophages (M?), stimulated M?(with interferon F or lipopolysaccharide), T cells, T blasts (stimulated with phytohemagglutinin and IL-2), B cells, proximaltubular epithelial cells (PTECs), human aorta endothelial cells (HAECs), and human umbilical vascular endothelial cells(HUVECs). C, Ficolin-2 protein binds to dying cells. Cell death was induced in Jurkat T cells by ultraviolet irradiation, andrecombinant Ficolin-2 was added for 1 hr. Cells were analyzed by flow cytometry with antibodies against Ficolin-2, AnnexinA5, and 7-AAD. Inlay shows Annexin A5 vs. 7-AADplot. The amount of Ficolin-2 binding to the cells is depicted for living cells(green line, R1), early apoptotic cells (pink line, R2), and late apoptotic cells (blue line, R3). D, Intragraft cytokine expressionin relation to Ficolin-2 genotype. mRNA expression levels of IL-1A, TNF>, and IL-6 were assessed in 57 pretransplantationbiopsies from donors with the Ficolin-2 wild-type variant and in 12 with the Ala258Ser variant.

482 www.transplantjournal.com Transplantation & Volume 94, Number 5, September 15, 2012

Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

surface patterns on target cells may be dominated by similaractions of dead-cell binding by MBL and C1q (12, 13, 37Y40).However, MBL is not expressed in the kidney (see Figure 2,SDC, http://links.lww.com/TP/A681) (41) or in monocytesand macrophages (data not shown). C1q staining in pre-transplantation biopsies is extremely scarce (42), whereas C1qmRNA expression levels in the kidney are comparable to thosefor Ficolin-2 (see Figure 2, SDC, http://links.lww.com/TP/A681).The immunohistochemical staining pattern and results from thecell-type study (Fig. 2A, B) point toward donor-derived myeloidcells expressing Ficolin-2. Such so-called passenger leukocytes arehighly immunogenic (43). The effect of HLA-DR matching isstrongest in the first 5 months after kidney transplantation, with apeak between 1 and 2 months, which may suggest that passengerleukocytes persist several months in the graft (44).

Ficolin-2 may effect clearance of dying cells throughN-acetylglucosamine (GlcNAc) recognition. GlcNAc is exposedon membranes of early apoptotic cells (45). Binding ofmacrophages to apoptotic cells and phagocytosis are inhib-ited when the macrophages are preincubated with GlcNAc(46, 47). Ficolin-2 may also be involved in macropinocytosis,which is a preferred mechanism of uptake of early apoptoticcells by differentiated macrophages (48). Macropinocytosis isinitiated by C1q and MBL after engagement with calreticulinand CD91 on macrophage cell surfaces (11). Ficolin-2 effi-ciently binds calreticulin and CD91 (13, 49), implying itsinvolvement in macropinocytosis. The increased bindingcapacity to GlcNAc of the Ficolin-2 Ala258Ser variant com-pared with the wild-type variant (50) may mean that Ficolin-2genotype affects opsonization efficiency of early apoptoticcells. Timely removal of dying cells leads to phagocytes withonly low capacity for T-cell stimulation and minimal pro-duction of proinflammatory cytokines (51, 52). The clinicalimpact of the donor Ficolin-2 genotype in kidney trans-plantation may be explained by a similar phenomenonoccurring within the graft.

In conclusion, the amino acid substituting polymor-phism Ala258Ser in exon 8 of the Ficolin-2 gene in the donorindependently predicts improved kidney graft outcome.Serum levels and tissue expression levels of Ficolin-2 are notinfluenced by genotype. Ficolin-2 protein, which binds todying cells, is present in donor kidneys in a passenger leukocyte-like pattern. Donor grafts with the Ficolin-2 Ala258Ser vari-ant contain significantly elevated expression of IL-6. TheAla258Ser Ficolin-2 variant binds with higher affinity toGlcNAc compared with the wild-type variant. We proposethat the Ala258Ser polymorphism in Ficolin-2, which isexpressed by phagocytozing cells in the donor kidney, leadsto a more efficient handling of injured cells, possibly re-sulting in decreased intragraft exposure to danger signalsand a dampened alloimmune response.

MATERIALS AND METHODS

Investigation of PRR Polymorphisms in Relationto Clinical Outcome

Recipients and DonorsPatients who underwent transplantation between 1995 and 2005 with a

kidney from a deceased, HLA-DRYcompatible donor in the Leiden Uni-

versity Medical Center (Leiden, the Netherlands) or in the Erasmus Medical

Center (Rotterdam, the Netherlands) were considered for investigation

(n=520). We studied SNPs in a randomly selected training set (n=189) and

verified results in a validation set (n=302). Demographics of the study

cohorts are shown in Table 1.

DNA and PolymorphismsDNA had been extracted from peripheral blood cells or splenocytes by the

Chelex procedure (53) or by classic salting out technique. With an in-house

developed system (54), 74 SNPs in 22 PRRs were investigated in DNA

from recipients and donors. The SNPs are localized in coding sequences or

promoter regions. Genotype frequencies are summarized in Supplemental

Table 1 (see Table 1, SDC, http://links.lww.com/TP/A681). Six DNA poly-

morphisms in the MBL2 gene were investigated in DNA from recipients only,

using the INNO-LiPA MBL2 line probe assay (Innogenetics, Gent, Belgium).

As quality control, duplicate samples at different time points from

34 persons were genotyped. A reproducibility of 99.75% was observed.

Polymerase Chain Reaction and Signal DetectionSNPs were investigated with a reverse dot probe assay (54), using in-

house developed primers and probes. DNA (È125 ng) was amplified in five

multiplex polymerase chain reaction (PCR) reactions and one single PCR

reaction, denatured, and hybridized to probes coated on nitrocellulose

membranes. Hybridization, wash steps, and visualization of the DNA

products were performed in Auto-LiPA 48 devices (Innogenetics). Hy-

bridized products were visualized on the basis of conversion of a chro-

mogen (NBT/BCIP) to a purple precipitate by biotin/streptavidin/alkaline

phosphatase complexes.

Clinical Outcome Parameters and Statistical ApproachGenotypes with a frequency less than 5% were left out of the statistical

analyses. Associations of SNPs with occurrence of acute rejection and ATG-

requiring (severe) rejection were tested. The SNPs were considered clinically

relevant only when an association with PG0.05 with the outcome parameters

were found in both the training and validation sets. Afterwards, a joint test

combining the samples from training and validation sets was performed,

adjusting for a false discovery rate at 5%.

Within the validation set, the particular SNPs were further analyzed by (1)

inclusion in a multivariate model with clinical risk factors, (2) inclusion in a

multivariate Cox regression model to test for independent prediction of graft

loss, and (3) testing for an association with death-censored graft loss, using

Kaplan-Meier curves and log-rank testing.

Further Investigation of Ficolin-2

SerumFicolin-2 levels were measured in serum from 36 patients with the

Ficolin-2 wild-type (GG) variant and from 39 patients with the Ala258Ser

(GT) variant, using a Ficolin-2 enzyme-linked immunosorbent assay kit

(Cell Sciences, Canton, MA). Sera had been obtained 44 T 34 days (range,

0Y146) before transplantation. Serum (kindly provided by Dr. Eric Claas,

Department of Medical Microbiology, Leiden University Medical Center,

the Netherlands) was investigated from 21 donors with the wild-type variant

and 11 donors with the Ala258Ser variant.

Pretransplantation Material and Cell TypesSixty-nine wedge biopsies before transplantation, not related to the initial

patient cohorts, were investigated for Ficolin-2 mRNA expression. RNA

extraction and quantitative PCR were performed as described previously (55,

56). Primers for Ficolin-2 were 5¶-CGA-GGC-GGA-GAA-GTA-CAA-TC-3¶

and 5¶-TGG-CAG-TTT-TTG-TAC-CAC-CA-3¶ and showed seven mis-

matches in total with the Ficolin-1 sequence. To check that expression

patterns for Ficolin-2 had not derived from Ficolin-1 expression, specificity

of the Ficolin-2 primers was tested by extending the probes with M13 tails

(5¶-GAC-GTT-GTA-AAA-CGA-CGG-CCA-GT-3¶) and by sequencing PCR

products with a DNA sequencer 3730xl Genetic analyzer (Applied Biosys-

tems, Carlsbad, CA). Furthermore, distinct expression patterns in RNA

* 2012 Lippincott Williams & Wilkins Eikmans et al. 483

Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

from different organs (liver, spleen, and kidney) (Cell Application, San

Diego, CA) and from PBMC fractions were observed for Ficolin-2 and Ficolin-

1 (see Figure 2, SDC, http://links.lww.com/TP/A681).

Ficolin-2 mRNA expression was investigated in the spleen, PBMC frac-

tions (n=9), cultured dendritic cells (granulocyte macrophageYcolony

stimulating factor and IL-4, without or with interferon (IFN)-F/IFN-A/

lipopolysaccharide stimulation for 6 hr), CD14+ monocytes (n=3), differ-

entiated macrophages (between 5 and 8 days of granulocyte macrophageYcolony stimulating factor or macrophage-colony stimulating factor, n=7),

stimulated macrophages (10-ng IFN-F or 100-ng lipopolysaccharide for

24 hr, n=6), T cells (n=5), T blasts (stimulated with phytohemagglutinin and

IL-2; n=5), CD19+ B cells (n=6), proximal tubular epithelial cells (n=2),

human aorta endothelial cells (n=5), and human umbilical vascular endo-

thelial cells (n=5).

Immunohistochemistry for Ficolin-2Protein staining for Ficolin-2 was carried out on tissue sections from four

different deceased-donor kidneys, which had been disapproved for trans-

plantation purposes. Sections (5 Km) of snap-frozen biopsies were air dried

and acetone fixed for 10 min. Ficolin-2 expression was assessed using a

digitonin-conjugated monoclonal antibody to Ficolin-2 (GN4; Hycult

Biotechnology, Uden, the Netherlands), followed by HRP-conjugated sheep

anti-DIG (Roche Diagnostics, Mannheim, Germany). After washing, sec-

tions were incubated with tyramide-fluorescein isothiocyanate in tyramide

buffer (NENTM Life Science Products, Boston, MA), washed, incubated with

HRP-labeled rabbit antifluorescein isothiocyanate (Dako, Glostrup, Den-

mark), and developed with 3¶3¶-diaminobenzidine (Sigma, St. Louis, MO).

Sections were counterstained with hematoxylin (Merck, Darmstadt, Germany)

and mounted with Imsol (Klinipath, Duiven, the Netherlands).

Binding of Ficolin-2 to Dying CellsLiving and apoptotic Jurkat T cells were incubated with Ficolin-2.

Experimental setup is described in the Supplemental Methods (see SDC,

http://links.lww.com/TP/A681).

Intragraft Cytokine Expression in Relationto Ficolin-2 Genotype

mRNA expression levels of IL-6, IL-1A, and tumor necrosis factor > were

assessed in 57 pretransplantation biopsies from donors with the wild-type

variant and in 12 with the Ala258Ser variant. Levels were normalized to

geometric mean signals of GAPDH and A-actin.

ACKNOWLEDGMENTSThe authors thank Nelleke Korfage, Orchid Sandifort,

Willem Verduyn, and Bernadette Boom for technical assis-tance; Eric van Zwet (Department of Medical Statistics andBioinformatics, Leiden University Medical Center, Leiden, theNetherlands) for additional advice on statistics; Karen Dixon(Department of Nephrology, Leiden University Medical Center,Leiden, the Netherlands); and Niels Rekers, Lloyd d’OrsognaSebastiaan Heidt, and Arend Mulder (Department of Immu-nohematology and Blood Transfusion, Leiden UniversityMedical Center, Leiden, the Netherlands) for providing celltypes for RNA analysis.

REFERENCES1. Jeannin P, Jaillon S, Delneste Y. Pattern recognition receptors in the

immune response against dying cells. Curr Opin Immunol 2008;20: 530.

2. Goldstein DR. Toll-like receptors and other links between innate andacquired alloimmunity. Curr Opin Immunol 2004; 16: 538.

3. Obhrai J, Goldstein DR. The role of toll-like receptors in solid organtransplantation. Transplantation 2006; 81: 497.

4. Bouwman LH, Roep BO, Roos A. Mannose-binding lectin: clinicalimplications for infection, transplantation, and autoimmunity. HumImmunol 2006; 67: 247.

5. Goldstein DR. Toll like receptors and acute allograft rejection. TransplImmunol 2006; 17: 11.

6. Alegre ML, Leemans J, Le Moine A, et al. The multiple facets of toll-likereceptors in transplantation biology. Transplantation 2008; 86: 1.

7. Goldstein DR. The identity of innate immune activators in organtransplantation: origins from within or exterior to the host? Am JTransplant 2007; 7: 1692.

8. Scaffidi P, Misteli T, Bianchi ME. Release of chromatin proteinHMGB1 by necrotic cells triggers inflammation. Nature 2002; 418: 191.

9. Leemans JC, Stokman G, Claessen N, et al. Renal-associated TLR2mediates ischemia/reperfusion injury in the kidney. J Clin Invest 2005;115: 2894.

10. Kruger B, Krick S, Dhillon N, et al. Donor Toll-like receptor 4 con-tributes to ischemia and reperfusion injury following human kidneytransplantation. Proc Natl Acad Sci U S A 2009; 106: 3390.

11. Ogden CA, deCathelineau A, Hoffmann PR, et al. C1q and mannosebinding lectin engagement of cell surface calreticulin and CD91initiates macropinocytosis and uptake of apoptotic cells. J Exp Med2001; 194: 781.

12. Nauta AJ, Raaschou-Jensen N, Roos A, et al. Mannose-binding lectinengagement with late apoptotic and necrotic cells. Eur J Immunol 2003;33: 2853.

13. Kuraya M, Ming Z, Liu X, et al. Specific binding of L-ficolin and H-ficolin to apoptotic cells leads to complement activation. Immuno-biology 2005; 209: 689.

14. Jensen ML, Honore C, Hummelshoj T, et al. Ficolin-2 recognizes DNAand participates in the clearance of dying host cells. Mol Immunol 2007;44: 856.

15. Palmer SM, Burch LH, Davis RD, et al. The role of innate immunity inacute allograft rejection after lung transplantation. Am J Respir CritCare Med 2003; 168: 628.

16. Ducloux D, Deschamps M, Yannaraki M, et al. Relevance of Toll-likereceptor-4 polymorphisms in renal transplantation. Kidney Int 2005;67: 2454.

17. Palmer SM, Burch LH, Mir S, et al. Donor polymorphisms in Toll-likereceptor-4 influence the development of rejection after renal trans-plantation. Clin Transplant 2006; 20: 30.

18. Berger SP, Roos A, Mallat MJ, et al. Association between mannose-binding lectin levels and graft survival in kidney transplantation. Am JTransplant 2005; 5: 1361.

19. Berger SP, Roos A, Mallat MJ, et al. Low pretransplantation mannose-binding lectin levels predict superior patient and graft survival aftersimultaneous pancreas-kidney transplantation. J Am Soc Nephrol 2007;18: 2416.

20. Tiberio L, Tiberio GA, Bardella L, et al. Mechanisms of interleukin-6protection against ischemia-reperfusion injury in rat liver. Cytokine2006; 34: 131.

21. de Vries DK, Lindeman JH, Tsikas D, et al. Early renal ischemia-reperfusion injury in humans is dominated by IL-6 release from theallograft. Am J Transplant 2009; 9: 1574.

22. Oetting WS, Schladt DP, Leduc RE, et al. Validation of single nucle-otide polymorphisms associated with acute rejection in kidneytransplant recipients using a large multi-center cohort. Transpl Int2011; 24: 1231.

23. Skol AD, Scott LJ, Abecasis GR, et al. Joint analysis is more efficientthan replication-based analysis for two-stage genome-wide associationstudies. Nat Genet 2006; 38: 209.

24. Hwang YH, Ro H, Choi I, et al. Impact of polymorphisms of TLR4/CD14 and TLR3 on acute rejection in kidney transplantation. Trans-plantation 2009; 88: 699.

25. Kruger B, Banas MC, Walberer A, et al. A comprehensive genotype-phenotype interaction of different Toll-like receptor variations in arenal transplant cohort. Clin Sci (Lond) 2010; 119: 535.

26. de Rooij BJ, van Hoek B, ten Hove WR, et al. Lectin complementpathway gene profile of donor and recipient determine the risk ofbacterial infections after orthotopic liver transplantation. Hepatology2010; 52: 1100.

27. Savill J, Dransfield I, Gregory C, et al. A blast from the past: clearanceof apoptotic cells regulates immune responses. Nat Rev Immunol 2002;2: 965.

484 www.transplantjournal.com Transplantation & Volume 94, Number 5, September 15, 2012

Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

28. Krysko DV, D’Herde K, Vandenabeele P. Clearance of apoptoticand necrotic cells and its immunological consequences. Apoptosis 2006;11: 1709.

29. Erwig LP, Henson PM. Clearance of apoptotic cells by phagocytes. CellDeath Differ 2008; 15: 243.

30. Kono H, Rock KL. How dying cells alert the immune system to danger.Nat Rev Immunol 2008; 8: 279.

31. Iyer SS, Pulskens WP, Sadler JJ, et al. Necrotic cells trigger a sterileinflammatory response through the Nlrp3 inflammasome. Proc NatlAcad Sci U S A 2009; 106: 20388.

32. Schumer M, Colombel MC, Sawczuk IS, et al. Morphologic, bio-chemical, and molecular evidence of apoptosis during the reperfusionphase after brief periods of renal ischemia. Am J Pathol 1992; 140: 831.

33. Supavekin S, Zhang W, Kucherlapati R, et al. Differential gene ex-pression following early renal ischemia/reperfusion. Kidney Int 2003;63: 1714.

34. Oberbauer R, Rohrmoser M, Regele H, et al. Apoptosis of tubularepithelial cells in donor kidney biopsies predicts early renal allograftfunction. J Am Soc Nephrol 1999; 10: 2006.

35. Toronyi E, Lord R, Bowen ID, et al. Renal tubular cell necrosis andapoptosis in transplanted kidneys. Cell Biol Int 2001; 25: 267.

36. Li J, Gong Q, Zhong S, et al. Neutralization of the extracellular HMGB1released by ischaemic damaged renal cells protects against renalischaemia-reperfusion injury. Nephrol Dial Transplant 2011; 26: 469.

37. Stuart LM, Takahashi K, Shi L, et al. Mannose-binding lectin-deficientmice display defective apoptotic cell clearance but no autoimmunephenotype. J Immunol 2005; 174: 3220.

38. Korb LC, Ahearn JM. C1q binds directly and specifically to surfaceblebs of apoptotic human keratinocytes: complement deficiency andsystemic lupus erythematosus revisited. J Immunol 1997; 158: 4525.

39. Navratil JS, Watkins SC, Wisnieski JJ, et al. The globular heads of C1qspecifically recognize surface blebs of apoptotic vascular endothelialcells. J Immunol 2001; 166: 3231.

40. Fraser DA, Laust AK, Nelson EL, et al. C1q differentially modulatesphagocytosis and cytokine responses during ingestion of apoptoticcells by human monocytes, macrophages, and dendritic cells. JImmunol 2009; 183: 6175.

41. Garred P, Honore C, Ma YJ, et al. MBL2, FCN1, FCN2 andFCN3VThe genes behind the initiation of the lectin pathway ofcomplement. Mol Immunol 2009; 46: 2737.

42. Sund S, Hovig T, Reisaeter AV, et al. Complement activation in earlyprotocol kidney graft biopsies after living-donor transplantation.Transplantation 2003; 75: 1204.

43. Lechler RI, Batchelor JR. Restoration of immunogenicity to passengercellYdepleted kidney allografts by the addition of donor strain dendriticcells. J Exp Med 1982; 155: 31.

44. Thorogood J, Persijn GG, Schreuder GM, et al. The effect of HLAmatching on kidney graft survival in separate posttransplantationintervals. Transplantation 1990; 50: 146.

45. Orlando KA, Pittman RN. Rho kinase regulates phagocytosis, surfaceexpression of GlcNAc, and Golgi fragmentation of apoptotic PC12cells. Exp Cell Res 2006; 312: 3298.

46. Pradhan D, Krahling S, Williamson P, et al. Multiple systems forrecognition of apoptotic lymphocytes by macrophages. Mol Biol Cell1997; 8: 767.

47. Duvall E, Wyllie AH, Morris RG. Macrophage recognition of cellsundergoing programmed cell death (apoptosis). Immunology 1985;56: 351.

48. Xu W, Roos A, Schlagwein N, et al. IL-10Yproducing macrophagespreferentially clear early apoptotic cells. Blood 2006; 107: 4930.

49. Duus K, Thielens NM, Lacroix M, et al. CD91 interacts with mannan-binding lectin (MBL) through the MBL-associated serine proteaseYbinding site. FEBS J 2010; 277: 4956.

50. Hummelshoj T, Munthe-Fog L, Madsen HO, et al. Polymorphisms inthe FCN2 gene determine serum variation and function of Ficolin-2.Hum Mol Genet 2005; 14: 1651.

51. Fadok VA, Bratton DL, Konowal A, et al. Macrophages that haveingested apoptotic cells in vitro inhibit proinflammatory cytokineproduction through autocrine/paracrine mechanisms involving TGF-beta, PGE2, and PAF. J Clin Invest 1998; 101: 890.

52. Ip WK, Lau YL. Distinct maturation of, but not migration between,human monocyteYderived dendritic cells upon ingestion of apoptoticcells of early or late phases. J Immunol 2004; 173: 189.

53. Verduyn W, Doxiadis II, Anholts J, et al. Biotinylated DRB sequen-ceYspecific oligonucleotides. Comparison to serologic HLA-DR typingof organ donors in eurotransplant. Hum Immunol 1993; 37: 59.

54. Van der Cruyssen B, Nuytinck L, Boullart L, et al. Polymorphisms inthe ficolin 1 gene (FCN1) are associated with susceptibility to thedevelopment of rheumatoid arthritis. Rheumatology (Oxford) 2007;46: 1792.

55. Heidt S, Roelen DL, Eijsink C, et al. Intravenous immunoglobulinpreparations have no direct effect on B cell proliferation and immu-noglobulin production. Clin Exp Immunol 2009; 158: 99.

56. D’orsogna LJ, Roelen DL, van der Meer-Prins EM, et al. Tissue specific-ity of cross-reactive allogeneic responses by EBV EBNA3AYspecificmemory T cells. Transplantation 2011; 91: 494.

* 2012 Lippincott Williams & Wilkins Eikmans et al. 485

Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.