Innate immune genetic profile to predict infection risk and outcome after liver transplant


  • Raymund R. Razonable

    Corresponding author
    1. Division of Infectious Diseases, Department of Medicine, and the William J. von Liebig Transplant Center College of Medicine Mayo Clinic Rochester, MN
    • Division of Infectious Diseases, Mayo Clinic, Marian Hall 5, 200 First Street SW, Rochester, MN 55905
    Search for more papers by this author
    • fax: 507-255-7767

  • Potential conflict of interest: Nothing to report.

  • See Article on Page 1100


MASP, MBL-associated serine protease; MBL, mannose-binding lectin; SNP, single-nucleotide polymorphism.

A deficiency in the plasma of an activator of complement has long been suspected as a factor that predisposes some individuals to recurrent infections. In 1968, a case of a 3-month-old girl with severe dermatitis, diarrhea, and recurrent bacterial infections was associated with a defective phagocytosis.1 Subsequent investigations demonstrated that the phagocytic defect was corrected in vitro by the addition of heterologous plasma, and thus, a deficiency in a plasma activator of complement was suspected. Because this was also observed in several of her relatives, a genetic origin was very likely.1 However, the exact nature of this plasma activator of complement was not known for 2 decades, until 1987, when this was suggested as mannose-binding lectin (MBL).2 Produced primarily by the liver, MBL is a major activator of the lectin pathway of complement, which serves a central role in innate immunity together with a sophisticated network of cells, barriers, and molecules that establish rapid, efficient, yet nonspecific defense against invading pathogens, while the pathogen-specific adaptive immune response is undergoing development.

MBL is one of the major soluble pattern recognition molecules that recognize pathogen-associated carbohydrate moieties present in microorganisms. MBL recognizes a variety of pathogens including gram-positive and gram-negative bacteria, fungi, viruses, and parasites, thereby highlighting its broad spectrum of immunologic activity. Three related soluble pattern recognition molecules that function to recognize acetyl groups (whether presented from carbohydrates or other compounds) are the ficolins (designated as ficolin-1, ficolin-2, and ficolin-3). Recognition of pathogen-derived polysaccharide and acetyl molecular patterns by MBL and ficolins, respectively, results in engagement of MBL-associated serine proteases (MASPs, primarily MASP-2),3 leading to complement activation (particularly the lectin pathway) and antimicrobial responses characterized by opsono-phagocytosis and formation of a membrane attack complex that causes direct lysis of the pathogen.

Serum concentrations of MBL and ficolins vary widely in humans. For example, serum MBL levels vary from as low as 5 ng to >10 μg/mL, although the level for each individual is quite stable throughout life. Serum MBL and ficolin concentrations are largely under genetic control. MBL is encoded by the MBL2 gene, which is located on chromosome 10. Three missense single-nucleotide polymorphisms (SNPs) in exon 1 of MBL2 (SNP at codons 52 [referred to as allele D], 54 [allele B], and 57 [allele C], which are all collectively called the O allele, whereas wild-type is referred to as A allele) and other SNPs in the promoter and 5′-untranslated region result in low serum levels of functional MBL (Table 1).4 The frequency of variant alleles vary in ethnic groups worldwide, and in some populations, the O allele may exceed a prevalence of 40%, and thus are relatively common in humans. Serum levels of L-ficolin, which is also produced primarily in the liver, is likewise influenced by genetic control (low levels have been associated with SNPs in the promoter region [at positions −986, −602, and −4] and in exon 8 [A258S]) of the FCN2 gene.

Table 1. MBL Genotypes, Haplotypes, and Phenotypes
HaplotypeDesignationMBL Levels
  1. Serum MBL levels vary substantially within each coding genotype, and polymorphisms in the promoter and the 5′-untranslated regions. Three missense SNPs in exon 1 of MBL2 (SNP at codon 52 [CGT {arginine} to TGT {cysteine}; referred to as allele D], codon 54 [GGC {glycine} to GAC {aspartate}; referred to as allele B], and codon 57 [GGA {glycine} to GAA {glutamic acid}; referred to as allele C]; all variants are collectively called the O allele, whereas the wild-type is referred to as the A allele) and other SNPs in the promoter region (position −550 [G to C; designated as alleles H and L, respectively] and position −221 [G to C; alleles X and Y] and at the 5′-untranslated region [C to T, designated as alleles P and Q, respectively]) are associated with low serum levels of functional MBL.4

LYAAIntermediate to High

Low levels of innate pattern recognition molecules may impair pathogen recognition, leading to an increased risk of various infectious diseases. This hypothesis has been investigated in several studies during the past decade.1-35 In this issue of HEPATOLOGY, de Rooij and colleagues investigated the hypothesis that functional SNPs within the lectin complement pathway (MBL2, ficolin-2 [FCN2], and the MASP2 gene) increases the risk of clinically-significant bacterial infections after liver transplantation.36 Using a sophisticated study design with principal and confirmatory cohorts, the authors observed three clinically relevant findings, namely (1) a stepwise increase in the risk of clinically significant bacterial infections with the number of SNPs in MBL2, FCN2, and MASP2; (2) an increase in the risk of infection if there was a mismatch between the donor and recipient in any of the genes (specifically when the wild-type recipient receives a liver allograft from a donor with variant genes); and (3) an increase in risk of death related to bacterial infections in patients who received liver from donors with variant genes.

These findings can have important implications in the clinical setting. Infections are common complications of liver transplantation, with many resulting in poor outcomes, despite the best efforts at prevention (with use of vaccination, antimicrobial prophylaxis, and preemptive therapy) and treatment. Genetic make-up has been suggested to play a role in infectious disease predisposition after liver transplantation, and the study by de Rooij and colleagues provides solid evidence of this.36 Indeed, there have been three prior studies that have consistently observed a higher incidence of clinically significant infections in liver transplant patients who received allograft from MBL-deficient donors (Table 2).7, 8, 35 It is also consistent among the four studies that the significant association was with the receipt of liver from a donor with variant genes.7, 8, 35, 36 This is not unexpected because these innate molecules (MBL and L-ficolin) are primarily produced in the liver, and in one of these studies,7 serum MBL levels declined consequent to transplantation with liver allograft from MBL2 variant donor.

Table 2. Studies on the Association Between Gene Polymorphisms in the Lectin Complement Pathway and Bacterial Infections After Liver Transplantation
Study (Author; Year)PopulationFindings
Bouwman et al. (2005)7n = 49MBL2 variant in the donor liver (but not recipient) was associated with a 3.8-fold higher incidence of clinically significant infections.
  Transplantation of MBL2 wild-type recipients with donor livers carrying the variant alleles resulted in rapid decline in serum MBL levels.
Cervera et al. (2009)8n = 95Receipt of liver from donor with MBL2 variant was associated with lower survival rate, and higher infection-related mortality (50% vs. 14%, P = 0.040).
  Patients who received liver from donor with MBL2 variant had higher incidence of bacterial septic shock (46% vs. 11%, P = 0.004).
Worthley et al. (2009)35n = 102MBL2 variant in the donor (but not recipient) was significantly associated with clinically significant infections (cumulative incidence: 55% in recipients of deficient livers vs. 32% for recipients of wild-type livers; P = 0.002).
  Low MBL levels were associated with clinically significant infections (cumulative incidence: 52% vs. 20%, P = 0.003).
  MBL2 mutation in the donor liver was independently associated with clinically significant infections (hazard ratio, 2.8; P = 0.005).
de Rooij et al. (2010)36Principal cohort (n = 143)In the principal cohort, recipients of a donor liver with polymorphisms in all three components (MBL2, FCN2, and MASP2) had a cumulative infection risk of 75% vs. 18% with wild-type donor livers (P = 0.002).
 Confirmatory cohort (n = 167)In the confirmatory cohort, the cumulative incidence of infections was higher in the presence of polymorphisms (56% with three variants, 26% with two variants, 15% with one variant, and 20% without genetic variants; P = 0.04).
  Patients with a lectin pathway gene polymorphism and infection had a six-fold higher mortality, of which 80% was infection-related.

Could we translate these findings clinically by using an innate immune genetic profile as a prognostic factor to assess the risk and outcome of infection after liver transplantation? Knowledge of this genetic profile can guide clinicians in individualizing patient management by tailoring the intensity of immunosuppression and the aggressiveness of microbial surveillance and prevention, depending on a specific genetic profile. Of interest, the authors also observed an increased risk of death from bacterial infections in patients with variant alleles, and thus, knowledge of this genetic profile may lead to early diagnosis of suspected infection and prompt initiation of antimicrobials to patients with variant alleles in an effort to prevent poor outcome. Spurred by the interesting observations reported in these four studies (Table 2), I anticipate more studies to be conducted involving a larger cohort of patients and ideally using a prospective study design. Meanwhile, although clinical studies in the transplant setting are emerging,7-9, 35, 36 I should recognize many other investigators who have pioneered investigations on the clinical implications of MBL deficiency in various immunocompetent and immunocompromised populations.2-6, 10-33, 36-38 Some studies have reported significant associations between low MBL or ficolin levels or genotypes with sepsis or systemic inflammatory response syndrome,14, 16, 33, 37 bacteremia and pneumonia,28, 34 postoperative infection,32 febrile neutropenia,27 and specific infections due to Streptococcus pneumoniae,30 Neisseria meningitidis,22 Pseudomonas aeruginosa and Burkholderia cepacia,18 and other invasive bacterial, viral, and fungal infections.20, 24-26 However, the associations have not always been consistent, with some studies demonstrating lack of significant association.6, 23 In contrast to MBL studies, the clinical implications of ficolin levels are less studied, although a recent case-control study demonstrated that low ficolin level was associated with susceptibility to gram-positive bacterial sepsis.31

It should be emphasized that although MBL deficiency states are common, the majority of affected humans are healthy and do not exhibit infectious disease predisposition or higher rates of mortality. This observation underscores the intricate nature of the immune system, which exhibits redundancy so that back-up defense mechanisms become operational when one pathway is defective. Hence, for a genotype to be clinically manifested, one may need to impair more than one aspect of the immune system. This theory could explain why the significant association between MBL deficiency and infections has been most pronounced in individuals with other immunocompromised states, such as neonates and young children with immature immune system, neutropenic patients due to cytotoxic chemotherapy, bone marrow transplant recipients, and in this study by de Rooij and colleagues,36 liver transplant recipients with pharmacologically impaired adaptive immune system.

Another major clinical implication of this study is the potential role of MBL replacement therapy in the prevention and treatment of infectious diseases, a novel avenue that is distinct from the traditional use of antimicrobial therapy. It is noteworthy to emphasize that the phagocytic defect of the 3-month-old child initially described in 1968 was corrected in vitro by heterologous plasma.1 Consequently, there has been remarkable interest in studying the clinical benefit of MBL replacement in deficient humans.38-40 Early phase clinical trials are underway. However, the phase 1 multicenter study aimed at evaluating the safety of recombinant human MBL after liver transplantation was terminated by the sponsor ( identifier NCT00415311). Currently, there is an ongoing phase 1 study to investigate the role of recombinant human MBL for treatment of MBL-deficient patients with fever and neutropenia ( identifier NCT00886496). The results of MBL replacement studies are eagerly anticipated.

In conclusion, there is an ever-increasing body of data to support the role of MBL, ficolins, and the lectin complement pathway as important components of the innate immune system. MBL deficiency states are common and may predispose individuals to a broad range of infectious diseases especially among individuals with immature, defective, or impaired adaptive immune systems, such as liver transplant recipients. The study of de Rooij et al. in this issue adds solid evidence on the role of immune gene deficiencies in the pathogenesis and predisposition to infectious disease after liver transplantation. The translation of these observations into clinical practice, either as a prognostic tool to assess risk, severity, and outcome of disease, is eagerly awaited. Likewise, the potential clinical utility of MBL supplementation in at-risk individuals is being actively pursued as a novel avenue for infection prevention and treatment.