The genetics of atypical hemolytic uremic syndrome

Atypical hemolytic uremic syndrome (aHUS) is a disorder characterized by thrombocytopenia and microangiopathic hemolytic anemia due to endothelial injury. aHUS is felt to be caused by defective complement regulation due to underlying genetic mutations in complement regulators or activators, most often of the alternative pathway. Mutations causing aHUS can be subdivided into two groups, loss of function mutations (affecting factor H, factor H-related proteins, membrane co-factor protein, and factor I), and gain of function mutations (affecting factor B and C3). As more information becomes available on the relationship between specific mutations and clinical outcome, complete genetic workup of aHUS patients becomes more and more important. In this review, we will discuss the genetic background of aHUS, the role of complement for aHUS pathogenesis, and the different groups of specific mutations known to be involved in the pathogenesis of aHUS.


Thrombotic microangiopathy
Thrombotic microangiopathy (TMA) defines a spectrum of disorders characterized by thrombocytopenia and microangiopathic hemolytic anemia (MAHA) due to endothelial injury. TMA was first described by Moschcowitz in 1924 [1][2][3] and is an ubiquitous disease that can involve multiple organs including the brain, heart, lungs, gastrointestinal tract, and kidneys [1]. In the two most common types of TMA, the major target organs are the central nervous system (CNS) in thrombotic thrombocytopenic purpura (TTP), and the kidneys in hemolytic uremic syndrome (HUS) [1]. In the majority of cases TTP is caused by IgG and IgM autoantibodies impairing the function of ADAMTS13, a disintegrin and metalloprotease with a thrombospondin type 1 motif, member 13 (i. e., Morbus Moschcowitz). In a small group of patients, TTP is caused by mutations in the gene encoding for ADAMTS13 (i. e., Upshaw-Schulman syndrome) [1,[3][4][5][6]. Both the acquired and genetic forms of TTP cause impaired ADAMST13 function and lead to a severely decreased (i. e., <10%) ADAMTS13 activity. Physiologically, ADAMTS13 specifically cleaves the ultra-large von Willebrand factor (VWF) multimers released from the vascular endothelium into smaller multimers, and ADAMTS13 deficiency results in the presence of ultra-large vWF in the circulation, leading to platelet adhesion and thrombus formation [1,3,4].
By contrast, HUS has various etiological factors. Most cases of HUS are caused by an infection with Shiga toxin (Stx)-producing Escherichia coli (STEC-HUS or eHUS) [1,[7][8][9]. After a prodromal phase of often bloody diarrhea, the Stx can enter the bloodstream and lead to damage of the vascular endothelium with subsequent thrombus formation and renal failure. Although most HUS cases occur in children and are the consequence of an infection with Stx-producing Escherichia coli (STEC), HUS can also be primarily caused by a dysregulation of the complement system (atypical HUS [aHUS]) or occur secondary to underlying conditions such as infections, drugs, pregnancy, and metabolic diseases (secondary TMA) [8,10,11].
In aHUS, genetic susceptibility and triggering events such as respiratory tract and gastrointestinal infections combine in the manifestation of a TMA event ("multiple-hit hypothesis") [5,8]. The fact that a first aHUS event can occur only later in life indicates the need for an additional trigger. aHUS is rare, with an incidence of approximately 0.5 million per year [12], can affect all age groups, runs in families, and comes with a significantly more severe phenotype than STEC-HUS. It is caused by defective complement regulation, specifically by mu-tations or autoantibodies affecting complement regulators or activators [5,8,10,11,13,14], most often of the alternative pathway [2].
Endothelial injury in aHUS patients ranges from only mild endothelial swelling to full occlusion of the vessel due to platelet adhesion and thrombus formation. Renal biopsies from aHUS cases show the classical pattern of TMA: glomerular capillary wall thickening, mesangiolysis, platelet/fibrin deposition, and narrowing of the vessels; on blood film typically (but not necessarily) erythrocyte fragments are seen [4,11]. Chronic TMA lesions, such as double contours of peripheral capillary walls, new subendothelial basement membrane, and subendothelial flocculent material, may be seen by electron microscopy [10].
The focus of this review will be on aHUS, the role of complement and the involvement of genetics in the pathogenesis of this rare disease.

The complement system
The complement system (. Fig. 1) is part of innate immunity. Its main functions are immune complex clearing, chemotaxis for recruiting inflammatory cells, opsonization of invading microbes, phagocytosis of foreign particles, and cell lysis [11,15]. The complement system consists of >30 proteins [4] and can be activated by three pathways: the clas-sical (CP), lectin (LP), and alternative pathway (AP). Although the classical and lectin pathways are inducible via immune complexes (classical) or the presence of certain carbohydrate sequences, for example, those found on bacterial surfaces (mannose-binding lectin [MBL] or ficolin to carbohydrate sequences), the alternative pathway is constitutively active and therefore requires tight regulation [5,8,[15][16][17]. All three activation pathways merge in the activation of complement C3 via the C3 convertase complex C3bBb (for the AP) or C4b2a (for the CP and LP) resulting in the formation of the anaphylatoxin C3a and the generation of the opsonin C3b. Via formation of C5 convertase (AP: C3bBbC3b or CP: C4bC2aC3b, respectively), C3b can induce the terminal complement pathway, resulting in the formation of the anaphylatoxin C5a and the membrane attack complex (MAC, C5b-9), causing cell lysis [5,8,[15][16][17].
Serum and membrane-anchored complement proteins combine to limit complement activation both temporally and spatially with factor H (FH) being the most important fluid phase regulator. FH belongs to the factor H family of proteins and consists of 20 short consensus repeats (SCRs). The N-terminus of FH with SCRs1-4 harbors the cofactor activity for factor I (FI)-mediated C3b cleavage (cofactor activity) and accelerates the natural decay of C3 convertases (decay accelerating activity). The C-terminus of FH with SCRs19-20 mediates the binding to C3b/C3d, heparin, glycosaminoglycans, and to the cell surface (. Fig. 2a; [5,6,11]). Structurally and possibly functionally similar to FH are the proteins of the group of factor H-related proteins (FHRs), including FHR1, FHR2, FHR3, FHR4, and FHR5 [18]. The two C-terminal SCR domains of FHR1-5 are (almost) identical to the FH C-terminus [18]. FHR1 lacks the regulatory function of FH; however, it has been shown to control the complement cascade at a later step by blocking C5 convertase activity and the formation of MAC/C5b-9 [18]. Since the coding region of the FHR genes is located next to the FH gene (i. e., chromosome 1q32) and carries a high rate of sequence ho-mology, it is highly susceptible for genetic rearrangements between FH and FHRs, leading to the formation of FH::FHR hybrid genes [7,11,14,18,19]. The most relevant hybrid genes for aHUS are composed of FH and FHR1 or FHR3 (. Fig. 3; [5,18]) and can lead to products with decreased local complement regulation on the endothelial cell surface [10,14,20].
The amount of activated complement is dynamically regulated, depends on multiple factors and varies per organ. The microenvironment of the kidneys with high blood flow, high concentrations of complement proteins, endothelial fenestrae with uncovered glomerular basement membrane and the local production of complement proteins likely results in a higher susceptibility of the kidney for complement-mediated insults.

Genetics
The main pathological feature of aHUS is endothelial injury. Over activation of the complement system caused by underlying genetic mutations, results in an increased level of MAC/C5b-9 deposition on the endothelial cell surface causing endothelial cell activation and damage [4,[7][8][9]11]. Endothelial activation and damage results in inflammation and activation of the coagulation cascade with thrombus formation as result (. Fig. 4; [8]). As genetic mutations predispose for complement dysregulation, screening of the genetic complement profile is important in aHUS patients.
The general tool widely used to detect genetic variants is sequence analy-sis of the coding regions of candidate genes in aHUS patients. This allows for a comprehensive analysis of all genetic variations in a candidate gene or set of candidate genes. With the development of next-generation sequencing (NGS) technologies, tens to hundreds of genes can be effectively analyzed simultaneously. However, whereas candidate gene sequencing is a targeted approach, whole exome sequencing (WES) or whole genome sequencing (WGS) can interrogate sequence variants in all coding regions/exons (WES) or even the entire genome (WGS). The sequence homology of some complement genes (for example, the genes from the FH operon) results in ambiguous read alignments from NGS data sets. For this reason, a diagnostic routing frequently consists of a combination of amplicon-based sequencing of the FH operon (at least SCR20 and the complete coding region of FHR1-5) and deep intronic single nucleotide polymorphisms (SNPs) of aHUS-associated haplotypes, in combination with an NGSbased gene panel aiming to establish sequence variants in aHUS patients. In the near future this will probably be replaced by NGS once the technology has become more cost effective.
In the human genome next to the FH gene, genes encoding FHRs1-5 are located in tandem next to FH on chromosome 1q31.3. All these genes originate from FH through gene duplication events [21]. The FH-FHRs loci contain several segmental duplications, making them prone to genetic structural rearrangements due to allelic and non-allelic homologous recombination events. This has led to copy number polymorphisms, for example, the common deletion that results in the loss of FHR1-FHR3 (Rs 6677604). In aHUS patients, genetic structural rearrangement in the FH operon may be observed, resulting in different FHR hybrid genes. For genetic diagnostics, this copy number variation in the FH-FHRs genomic region can be evaluated using multiplex ligation-dependent probe amplification (MLPA) [22]. In the future, novel longread NGS technology may replace (part of) this MLPA approach.

Interpretation of genetic results
Functional genomics will become more and more important in unraveling the functional consequences of genetic aberrations. The likely functional impact of identified sequence variations can be predicted using in silico software (such as PolyPhen, Align GVD, and SIFT), but these are not always accurate. Several established mutations identified in aHUS patients have dbSNP numbers (numbers of the Short Genetic Variations database), as they have been identified in large NGS projects (for example, the p.Arg1210Cys and rs12191309). For most functional analysis methods that are available today (such as crystallography, Biacore, hemolytic assays) and/or cofactor assays, recombinant protein production and purification or isolation of mutant proteins from patients' blood is needed, which may be a time-consuming and challenging matter. Because of its complexity, it should be performed in centers of expertise.

Specific genetic mutations in aHUS
The first report on a patient with aHUS was published in 1981 [23], when Thompson and Winterborn described an 8-month-old Asian boy presenting with symptoms of HUS and very low C3 levels but normal C4 levels [23]. This case was published before the era of genetic screening, and it was only later that Warwicker et al. (1998) linked a case of familial aHUS with a FH mutation [24]. The discovery of mutations of several other genes, encoding for complement-activating or -regulating proteins, followed (. Fig. 5).
Mutations causing aHUS can be subdivided into loss of function mutations of complement regulators (FH, FHRs, MCP/CD46, FI), or gain of function mutations of complement activators (FB, C3). In addition, FH autoantibodies can be identified, which in most patients coincide with the presence of a homozygous deletion in FHR1 and/or FHR3 [2, 6-8, 10, 11, 14-16, 25-27]. The genetics of atypical hemolytic uremic syndrome Abstract Atypical hemolytic uremic syndrome (aHUS) is a disorder characterized by thrombocytopenia and microangiopathic hemolytic anemia due to endothelial injury. aHUS is felt to be caused by defective complement regulation due to underlying genetic mutations in complement regulators or activators, most often of the alternative pathway. Mutations causing aHUS can be subdivided into two groups, loss of function mutations (affecting factor H, factor H-related proteins, membrane co-factor protein, and factor I), and gain of function mutations (affecting factor B and C3). As more information becomes available on the relationship between specific mutations and clinical outcome, complete genetic workup of aHUS patients becomes more and more important. In this review, we will discuss the genetic background of aHUS, the role of complement for aHUS pathogenesis, and the different groups of specific mutations known to be involved in the pathogenesis of aHUS.

Schlüsselwörter
Alternativer Komplementweg · Komplementsystem · Endothel · Mikroangiopathische haemolytische Anaemie · Thrombotische Mikroangiopathie ment regulators and cause ongoing excessive complement activation with subsequent endothelial cell injury. Mutations involved affect the genes encoding for FH, FHRs, MCP/CD46, and FI [8,14]. Although these mutations lead to uncontrolled activation of complement proteins on surfaces such as the vascular endothelium, complement regulation in the fluid phase remains intact in most cases. The variation in severity of complement regulation may be one explanation for the normal systemic C3 levels seen in many of aHUS patients. Thus, normal plasma C3 levels can be found in aHUS patients, and do not rule out the presence of genetic abnormalities.
Studies have shown that some aHUS patients carry one or more of the defined risk haplotypes (mostly SNPs) of the genes encoding for FH, FHR1 or MCP/CD46, in addition to a complement mutation [4,13,17,32,33]. In combination with a genetic mutation, these risk haplotypes lead to an increased risk for the development of aHUS, but a haplotype itself is seen as a susceptibility factor rather than the direct cause of the disease [13,17,32,33].
As mentioned above, aHUS can also be due to FH::FHR hybrid genes, which are caused by rearrangements between FH and FHR genes. There are only a few such cases described in the literature, and the incidence of hybrid genes in aHUS is estimated to be around 1-5% [7,10]. The first case was published in 2006 by Venables et al. [34]. They described a case of a family with multiple members with aHUS, and used cDNA analysis for genetic screening. This family screened positive for an FH::FHR1 hybrid gene [34]. In 2010, Noris

Gain of function mutations
Gain of function mutations are mutations in the complement proteins FB and C3. Gain of function mutations lead to enhanced stability and thus activity of the alternative C3 convertase, which is resistant to regulation, resulting in the over-production of the terminal complement pathway product MAC/C5b-9 [14].
Analysis of the Global aHUS Registry (2018) showed that 4 out of 275 patients (2%) had a mutation in FB, and 21 out of 331 (6%) had a mutation in C3 [12].

Noncomplement-related genetic mutations in aHUS
Because of the extended use of genetic testing toanalyze large patientcohorts, on rare occasions new genetic mutations are found. In a small group of patients with the clinical diagnosis aHUS, a rare mutation is found in non-complement-related proteins. Examples are diacylglycerol kinase epsilon (DGKE), inverted formin-2 (INF2), and plasminogen (PLG).
DGKE is a protein involved in cell metabolism, and is responsible for the catalyzation of diacylglycerol to phosphatidic acid, both lipids playing an important role in cell lipid signaling. DGKE is also responsible for the modulation of protein kinase C (PKC) activity [8,14]. DGKE mutations result in the activation of PKC, which stimulates pro-thrombotic factors such as VWF and tissue factor [3,42]. Patients with homozygous DGKE mutations typically present with a TMA episode during the first year of life, develop proteinuria, and progress to end stage renal disease during their second to third decade of life [3,8,9,42]. DGKE was first described by Lemaire et al. in 2013, who found mutations in 9 patients using exome sequencing [42]. Mutations in DGKE are rare, only a few cases have been published [4], and a link to com-   FH factor H, MCP/CD46 membrane cofactor protein, FI factor I, FB factor B, FHR factor H-related proteins, THBD/CD141 thrombomodulin, aHUS atypical hemolytic uremic syndrome, ESRD end-stage renal disease plement has not been identified [2-4, 8, 9, 43]. INF-2 belongs to a family of proteins called formins. It plays an important role in the regulation of certain cytoskeletal functions as it accelerates actin polymerization and depolymerization [44]. Challis et al. [41] used WES to analyze 28 families with cases of aHUS and used Sanger sequencing to analyze 161 sporadic aHUS cases. In 2 families, they identified mutations in the gene encoding for INF2 [44], whereas no INF2 mutations were found in the sporadic aHUS cases. Functional analysis of the mutations showed modified INF2 localization and disrupted cytoskeleton [44]. Members of one of the families carried a homozygous FH haplotype and was homozygous for a MCP/CD46 risk haplotype, whereas members of the other familycarried a homozygous MCP/CD46 risk haplotype and one copy of a FH haplotype [44].
PLG is another protein found to be related to aHUS, but not related to the complement cascade. PLG is encoded by the PLG gene [22] and becomes converted into plasmin during clot formation. Plasmin is important for the resolu-tion of blood clots and thus the regulation of the coagulation pathway. Decreased production of PLG results in decreased plasmin levels with diminished dissolution of blood clots and dysregulation of the coagulation pathway, eventually resulting in thrombus formation. Bu et al. [22] used targeted genomic enrichment and massive parallel sequencing to screen 36 patients with aHUS. They found 4 patients with a variant in the PLG gene [22] and concluded that those variants were related to the disease course of aHUS.

Clinical utility of genetic information
Published data have shown and described that the different genetic mutations known to be involved in aHUS come with a different age at disease onset, phenotype/genotype relationship, and risk of recurrence (. Table 1; [5, 7, 10-14, 25, 28, 30, 31, 33, 36, 43, 45]). Functional studies of diseasespecific genetic variants will help to identify pathogenic mechanisms and guide development of focused therapies. Knowledge of the pathological implications of complement genetic makeup will allow for an individualized assessment of disease predisposition, facilitating the implementation of personalized and preventive medical procedures and personalized treatment plans.

Conclusion
In this review, we have described the currently known mutations involved in aHUS pathogenesis. As more information becomes available about the relationship between specific mutations and the course of disease, complete genetic workup becomes increasingly important in patients with aHUS. Eventually, the underlying genetic mutation will play an important role in the choice of personalized treatment options. This article does not contain any studies with human participants or animals performed by any of the authors.

Abbreviations
Open Access. Thisarticleisdistributedundertheterms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/ 4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.