A review of clinical characteristics and genetic backgrounds in Alport syndrome

Alport syndrome (AS) is a progressive hereditary renal disease that is characterized by sensorineural hearing loss and ocular abnormalities. It is divided into three modes of inheritance, namely, X-linked Alport syndrome (XLAS), autosomal recessive AS (ARAS), and autosomal dominant AS (ADAS). XLAS is caused by pathogenic variants in COL4A5, while ADAS and ARAS are caused by those in COL4A3/COL4A4. Diagnosis is conventionally made pathologically, but recent advances in comprehensive genetic analysis have enabled genetic testing to be performed for the diagnosis of AS as first-line diagnosis. Because of these advances, substantial information about the genetics of AS has been obtained and the genetic background of this disease has been revealed, including genotype–phenotype correlations and mechanisms of onset in some male XLAS cases that lead to milder phenotypes of late-onset end-stage renal disease (ESRD). There is currently no radical therapy for AS and treatment is only performed to delay progression to ESRD using nephron-protective drugs. Angiotensin-converting enzyme inhibitors can remarkably delay the development of ESRD. Recently, some new drugs for this disease have entered clinical trials or been developed in laboratories. In this article, we review the diagnostic strategy, genotype–phenotype correlation, mechanisms of onset of milder phenotypes, and treatment of AS, among others.


Introduction
Alport syndrome (AS) is a progressive hereditary renal disease accompanied by sensorineural hearing loss and ocular abnormalities. AS develops because of pathogenic variants in the COL4A3, COL4A4, and COL4A5 genes encoding type IV collagen α3, α4, and α5 chains, respectively [1][2][3]. These type IV collagens constitute the glomerular basement membrane (GBM). AS has been reported to develop in approximately one in 5000 people; it comprises 0.5% of newly developed end-stage renal disease (ESRD) cases in adults [4] and 12.9% in children [5]. Type IV collagen has six different α chains, α1 to α6, which construct triple helix structures in which the three chains are combined. The combination of three α-chains is organ-specific: in the GBM, cochlea basement membrane, and base of the ocular lens, the triplet α3-α4-α5 is present, while in Bowman's capsule and skin basement membrane, it is α5-α5-α6. When an abnormality occurs in the α-chain, these triple helix structures are broken, causing nephropathy, sensorineural hearing loss, and eye lesions. AS is divided into X-linked Alport syndrome (XLAS), autosomal recessive AS (ARAS), and autosomal dominant AS (ADAS), according to its mode of inheritance. The distribution of these cases is reported to be as follows: 80% XLAS, 15% ARAS, and 5% ADAS. XLAS is caused by abnormality of COL4A5, while ADAS and ARAS are caused by abnormality in COL4A3 or COL4A4 [3]. Table 1 shows the diagnostic features established by the members of the Working Group for Alport Syndrome in the Japanese Society of Pediatric Nephrology (JSPN) in 2015 [6]. The main criterion is persistent hematuria. When patients fulfill one or more secondary features, or two or more accessory features, in addition to the primary feature, they can be diagnosed with AS. Symptoms differ depending on the mode of inheritance, with their features described below.

1) XLAS
In XLAS, there is often a family history of hematuria (with or without proteinuria) or renal failure. However, in approximately 15% of cases, there are de novo variants without a family history [3]. Microscopic hematuria is observed in all male cases. In female patients, it is observed in approximately 98% with hematuria and Table 1 Diagnostic features of Alport syndrome (revised in February 2015) prepared by the Working Group on Alport Syndrome of the Japanese Society of Pediatric Nephrology In addition to the primary feature, patients satisfy one or more secondary features If there is no corresponding item in secondary features, patients should satisfy two or more of the accessory features If patients have only the primary feature and have a family member diagnosed with Alport syndrome, the case is set as a "suspected case" Asymptomatic carriers can be diagnosed with any one feature of type IV collagen (II-1 or II-2) among the secondary features For all features, those caused by other diseases should be excluded, for example, a family history of kidney failure due to diabetes or senile deafness *1 Persisted for more than 3 months, which was confirmed by urinalysis on at least two occasions. As a rare situation, there is a possibility that hematuria may disappear at the time when renal failure progresses to the end stage, in which case appropriate examination such as of kidney dysfunction should be performed *2 This refers to a homozygous or heterozygous mutation of COL4A3 or COL4A4, or a hemizygous (male) or heterozygous (female) mutation of the COL4A5 gene *3 Type IV collagen α5 chain exists not only in the glomerular basement membrane and Bowman's capsule, but also in the skin basement membrane. Upon immunostaining using anti-α5-chain antibody, normal glomeruli and skin basement membrane are stained linearly and continuously. However, glomeruli, Bowman's capsule, and the skin basement membrane of male patients with X-linked Alport syndrome are completely negative. In glomeruli, Bowman's capsule and skin basement membrane of female patients are partially stained. In autosomal recessive Alport syndrome, the α3-, α4-, and α5 chains are not stained in glomerular basement membranes, whereas in Bowman's capsule and skin, normal α5-chain staining is shown. Note that the above is a typical pattern, but an atypical pattern also exists. Moreover, Alport syndrome cannot be ruled out even if α5 chain expression shows completely normal pattern *4 Glomerular basement membrane-specific abnormalities include broad irregular thickening of the glomerular basement membrane and reticulation of the lamina densa. Extensive thinning of the glomerular basement membrane frequently seen in benign familial hematuria is also seen in Alport syndrome, which can be the only finding of the glomerular basement membrane. In these cases, there is a high possibility of Alport syndrome if the cases show hearing loss, ocular abnormalities, or a family history of renal failure *5 Specific ocular abnormalities include anterior lenticonus, posterior subcapsular cataract, posterior polymorphous dystrophy, and retinal flecks  [7]. In male patients, proteinuria is recognized at an early stage of childhood, sometimes exhibiting a nephrotic status; it is also reported that 90% of patients develop ESRD by the age of 40 years, with the median age of development of ESRD being 25 years [8]. In females, 12% of cases have developed ESRD by the age of 40 [9]. We recently reported that patients exhibited proteinuria at a median age of 7 years, and that ESRD development occurred at a median of 65 years in female XLAS [7].
Sensorineural hearing loss often occurs from the latter stage of childhood; 90% of male patients and approximately 12% of female patients present with hearing loss by the age of 40 years [8,9]. Jais et al. reported that, in females, the presence of hearing loss predicts the development of ESRD; however, our recent study showed no significant difference in kidney severity between patients with and without hearing loss [7,9]. Specific ocular abnormalities, including anterior lenticonus, posterior subcapsular cataract, posterior polymorphous dystrophy, and retinal flecks, are known as ophthalmologic complications; clinical symptoms, such as obvious visual impairment, rarely appear [10]. Other complications include leiomyoma due to contiguous gene deletion encompassing the 5′ end of COL4A5 and COL4A6 intron 2. In this phenotype, leiomyoma is observed with full penetration, for which no gender difference is observed [11].

2) ARAS
Clinically, ARAS shows symptoms similar to those in male XLAS patients. There are no gender differences in clinical symptoms and incidence, and this condition sporadically occurs in one generation. Families with monoallelic variant carriers are often asymptomatic or show only microscopic hematuria (and mild proteinuria) [12,13]. For the genetic diagnosis of ARAS, analysis of at least one familial member (ideally, both parents) is necessary to prove that two heterozygous variants are located in trans positions on two different alleles (either COL4A3 or COL4A4). In terms of the clinical findings that we previously reported, the median age of ESRD development is 21 years. Sensorineural deafness was observed at a median age of onset of 20 years [12].

3) ADAS
Recently, we published an article regarding the clinical picture, pathology, and genetic background of ADAS [14]. The median age for detecting proteinuria was 17.0 years, and that for developing renal insufficiency was 70 years. In addition, both hearing loss and eye lesion were reported to occur quite rarely. Moreover, three of 16 patients with pathological findings showed focal segmental glomerular sclerosis (FSGS), as revealed by light microscopy. It has also been reported that, in approximately 10% of patients with familial focal segmental glomerulosclerosis, COL4A3 or COL4A4 mutations are identified, suggesting that there are many undiagnosed ADAS patients [15].

Pathological findings
There are no specific light microscopic findings in AS; nonspecific findings are observed, such as mesangial proliferation, FSGS, renal tubular atrophy, foam cell formation, or interstitial fibrosis. Electron microscopic (EM) findings show irregular thickening and thinning of the GBM, and lamellation and splitting in lamina densa can be observed (Fig. 1). These findings are specific to AS and electron microscopic findings are indispensable for the pathological diagnosis of this condition. However, pathological findings become apparent in the form of nephritis progression, even in male XLAS cases or ARAS cases; for female XLAS and ADAS cases, typical findings of EM can be observed generally at later stages and often show only thinning of GBM. Therefore, careful observation is necessary, because there are cases with no obvious changes except for thin basement membrane (TBM) at the early stages of AS. Immunohistologically, specific findings can be identified with α5 staining. Approximately 80% of XLAS male patients are completely negative for α5 staining, while XLAS female patients exhibit a mosaic pattern due to the mechanisms of X-chromosome inactivation that occur in female cells (Fig. 2) [2,16]. In ARAS patients, the GBM is negative for α3, α4, and α5 staining, while Bowman's capsule is α5 positive; this is because the GBM consists of α3-α4-α5, but Bowman's capsule consists of α5-α5-α6 and COL4A3/COL4A4 mutations do not influence α5 expression in Bowman's capsule. In ADAS patients, both the GBM and Bowman's capsule exhibit a normal pattern of α5 expression. α5 staining is a useful diagnostic method, because it remains unchanged regardless of age; however, it should solely be used as a method of auxiliary diagnosis, because it can show nonspecific findings, such as a normal expression pattern in more than 20% of XLAS males [1]. We have reported that 29% of male XLAS and 20% of ARAS cases showed atypical positive expression of α5 in GBM [12,17]. All male XLAS cases with α5 positivity possessed non-truncating variants or somatic mosaic variants in COL4A5 and showed significantly milder phenotypes of later-onset ESRD [17]. ARAS cases with α5 positivity possessed a missense variant in at least one allele in the COL4A3 or COL4A4 gene and again showed a milder phenotype that involved later development of ESRD [12].
Skin basement membrane consists of type IV collagen α5-α5-α6 triple helix. Therefore, α5 staining is a useful auxiliary diagnostic method only for XLAS cases: male Autosomal recessive Alport syndrome case shows negative expression only on GBM and positivity on BC, because BC consists of the α5-α5-α6 triple helix. E Schema for X-chromosome inactivation (XCI). In all female cells, either of the two X chromosomes is randomly inactivated. When the wild-type chromosome has been inactivated, the cell will not produce α5. Then, GBM and BC will be stained with a mosaic pattern  (Fig. 3A, B) [2]. It should be noted that decreased α5 expression can be observed at the bottom of the papillary epidermal basement membrane in normal skin (Fig. 3C, D) [18].

Gene tests
Along with the progression of gene analysis technology in recent years, there has been remarkable development of the genetic diagnosis system for AS. Because all three of the COL4A3, COL4A4, and COL4A5 genes consist of approximately 50 exons, conventional Sanger sequencing is too laborious and time-consuming for genetic screening. Therefore, the main strategy for gene screening in AS has become targeted sequencing, which captures all exons and exon-intron boundaries of all three genes, along with comprehensive sequencing analysis by NGS. Genes responsible for kidney diseases that clinically or pathologically resemble AS should be included in the targeted screening panel.
For example, BOR syndrome shows nephritis and deafness; patients with these symptoms are sometimes suspected of having AS. Moreover, the pathological findings of Pierson syndrome frequently include lamellation of the GBM, as revealed by EM. Moreover, some cases of Pierson syndrome with missense variants in the LAMB2 gene tend to show milder phenotypes involving later onset of ESRD [19]. These clinical features resemble those of AS. Cases with gene variants in LMX1B (nail-patella syndrome), PAX2 (renal coloboma syndrome), or MYH9 (MYH9 nephropathy) also show lamellation of the GBM, so they should be included in the gene list. Most mutations detected in AS are novel variants, because there are no mutational hotspots among the three genes. When we fail to detect pathogenic variants by means of targeted sequencing, we must proceed to the next step of detecting copy number variations (CNVs) and deep intronic pathogenic variants. Our gene screening strategy for AS is as follows: Step 1: targeted next-generation sequencing with a custom disease panel; when pathogenic variants are detected, they are confirmed by Sanger sequencing. When we fail to detect mutations by Step 1, we proceed to Step 2, which involves conducting pair analysis comparing NGS data from patients and normal controls, to screen for CNVs. When pair analysis shows the possibility of CNVs in any of the three genes, we proceed to detecting CNVs by multiplex ligation-dependent probe amplification (MLPA). When we fail to detect pathogenic variants with Steps 1 and 2, we proceed to Step 3: reverse transcription-polymerase chain reaction (RT-PCR) of mRNA and direct sequencing to detect aberrant splicing due to intronic or exonic variants. With these three steps, we can detect pathogenic variants in any one gene for more than 90% of cases clinically suspected of being AS [7,12,14,20]. With Step 2, we recently detected CNVs in COL4A5 in AS cases [21]. Pair analysis also revealed a case with clinically suspected AS because of ESRD and hearing loss, which had CNVs in the EYA1 gene, and diagnosed this as BOR syndrome [21]. With Step 3, we have identified deep intronic variants that lead to cryptic exon insertion by creating novel splicing sites [12,20]. We have also identified aberrant splicing caused by a so-called "silent variant," an exonic single-nucleotide substitution that does not change the coding amino acid [22].

Genotype-phenotype correlation
Three reports have been published analyzing the correlation between genotype and clinical picture based on large-scale data in male XLAS patients; we briefly summarize them here.  (4) 22 years with both large-and small-deletion mutations. Furthermore, the position of mutations closer to the 5′ end tended to show earlier progression of ESRD; in the same analysis, limited to only missense mutations in glycine, there was no correlation between location and severity. In these respects, the results differed from those in the report by Gross et al. [24].
Regarding female XLAS cases, two previous reports showed no genotype-phenotype correlations [7,9]. It has long been suspected that an uneven pattern of X-chromosome inactivation (i.e., skewed X chromosome inactivation) would influence the severity of XLAS in females [25]; however, no study has systematically proven the correlation between a skewed X pattern and clinical severity in female XLAS. In our experience, X-chromosome inactivation pattern analysis cannot predict kidney prognosis in female XLAS cases (manuscript in preparation).
For ARAS, two previous reports discussed the genotype-phenotype correlation in this condition. In ADAS, no genotype-phenotype correlations have been observed thus far. Even within one family, clinical severity differed significantly and some individuals developed ESRD, whereas others possessing the same mutation showed no urinary abnormality [14].

Mechanisms of onset of atypical mild phenotypes in male XLAS
It has been reported that some male XLAS cases show atypical mild phenotypes of late-onset ESRD. Based on previously reported results, it has been clarified that, for the following reasons, XLAS male patients may exhibit a mild clinical picture.

Cases with missense mutations
In all large-scale studies that analyzed genotype-phenotype correlations, it was reported that the age at which ESRD developed was significantly delayed in cases possessing missense mutations, as described above [8,23,24]. We have reported a male patient with very mild XLAS who had a missense mutation (p.Gly1000Val) in the COL4A5 gene and only showed hematuria at the age of 38 years [26]. We have also encountered a 40-year-old male with the same mutation who only showed hematuria (unpublished data). Thus, some missense mutations can lead to extremely mild phenotypes.

Cases with in-frame deletion mutations
For mutations involving deletion of a multiple of three nucleotides, the fact that three bases code one amino acid means that this mutation does not significantly influence the subsequent amino acid sequence (i.e., no frameshift occurs). Such mutations are called in-frame deletion mutations. Gross et al. reported that, when an in-frame mutation occurs, the age at which ESRD develops is significantly later than in cases with frameshift mutations [23]. In addition, Jais et al. classified in-frame deletions with deleted nucleotides numbering from 3 to 18 bases into a missense mutation group and analyzed them, because they showed mild phenotypes [8]. We also have reported two families with mild disease phenotypes, the members of which progressed to end-stage renal failure in their 60 s and 40 s due to deletions of 9 and 36 bases, respectively [17].

Cases with splice site mutations result in in-frame deletion at the transcript level
Furthermore, we recently reported a male with a deletion of 105 bases at the mRNA level due to a splice site mutation, in whom ESRD had not developed at the age of 47 [20]. This case showed an in-frame mutation at the mRNA level, indicating that mutation at a splice site may be associated with a mild clinical picture. Recently, we published the results of a large data analysis comparing splice site mutations, with or without in-frame deletion, at the transcript levels; we revealed that renal prognosis differs significantly, and that ESRD developed 9 years later in the in-frame deletion group [27].

Somatic cell mosaicism in proband cases
Mosaic mutations involve the presence of two cell populations with different genotypes in one individual, which have developed from a single fertilized egg (Fig. 4). Probands in cases of hereditary diseases can sometimes develop to retain mutants at somatic cell division failure. When this gene mutation is inserted after repeated cell division in fertilized eggs, cells with a normal gene and cells with an abnormal gene coexist; this is known as mosaicism. When this status is observed among somatic cells, it is recognized as somatic mosaicism. We previously reported a male patient with mild XLAS who had a mutation with somatic mosaicism [28]. Recently, we reported a tendency for an association between variant allele frequency and severity of renal symptoms in four men with XLAS with somatic mosaic variants [29]. Variant allele frequency was analyzed by ultra-deep sequencing with NGS. One of the cases, whose two daughters exhibited typical findings of female XLAS, was completely asymptomatic.

5.
A case with splice site mutation, in which both normal mRNA and aberrantly spliced mRNA are produced in renal tissue during transcription With splice site mutations, exon skipping is usually observed. However, incomplete mutation of the splice site may produce both normal and abnormal kidney mRNA, leading to milder phenotypes. This mechanism has been reported in another inherited disease, galactosialidosis [30]. In this disease, the IVS8 + 9C > G variant in the CTSA gene produces abnormally spliced mRNA with a low level of mutaƟon somaƟc mosaicism germline mosaicism somaƟc/germline mosaicism somaƟc cells germ cells Fig. 4 Schema of genetic mosaicism. When gene mutation occurs after repeated cycles of cell division in fertilized eggs, cells with a normal gene and cells with an abnormal gene coexist; this is known as mosaicism. When this status occurs in somatic cells, it is recognized as somatic mosaicism; in gonadal cells, it is known as germline mosaicism normal mRNA, which leads to a milder phenotype in the patient. A similar mechanism was also observed in a male XLAS case with a mild phenotype. We recently reported a mild XLAS case that exhibited only hematuria and very mild proteinuria despite being 22 years; the patient possessed a deep intronic mutation that produced both normal and abnormal mRNA in the kidney [20].

Cases with α5 expression in GBM
In a typical XLAS male patient, because of abnormality of the COL4A5 gene, expression of α5 is not recognized at all in GBM. However, the expression of α5 was confirmed in 29% of male XLAS patients; in those α5-positive cases, it was found that the age at which urinary protein was exhibited was significantly older, the amount of urine protein was significantly lower, the age of end-stage renal failure was also significantly older (37 years vs. 24 years), and the rate of hearing loss was also significantly lower. Thus, it was shown that α5-positive cases may exhibit a mild clinical picture [17]. Similar results were also recently published by a different group [31]. These cases with α5-positive expression in GBM possessed non-truncating (missense or in-frame deletion) mutations or somatic mosaic mutations in COL4A5 [17].

Treatment
There is currently no radical therapy for AS; treatment is only performed for the purpose of delaying progression to renal failure using nephroprotective drugs. A randomized controlled trial (RCT) is underway in Germany (EARLY PRO-TECT ALPORT study) for confirming nephroprotective effects. Two RCT trial results have been previously published examining reduction of urine protein levels by angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin receptor blockers (ARBs) in AS [32,33]. One compared losartan (ARB) and placebo, while another compared enalapril (ACEI) and losartan; both RCTs showed urine protein-reducing effects in both ACEI and ARB. A large retrospective study reported that ACEIs have the effect of delaying the progression to ESRD in AS [34]. Recently published expert guidelines recommend that male patients initiate treatment with ACEI from the time of diagnosis with AS [35]. Meanwhile, in female cases, it is recommended to start treatment at the time when urinary protein is detected, in addition to urine occult blood [35]. For ARAS, ACEI should be started at the time of diagnosis regardless of gender and, for ADAS, it should be initiated at the time when urinary protein is detected. No clinical trial has been performed comparing single (ACEI or ARB) and double block (ACEI and ARB); theoretically, double block can show a stronger renoprotective effect, and is already applied in clinical settings. It was previously reported that cyclosporine treatment for AS significantly reduced urine protein levels and showed renoprotective effects in long-term follow-up [36,37]. However, three recent reports showed that, although cyclosporine had strong effects for reducing urine protein, it accelerated interstitial fibrosis and had no renoprotective effects [38][39][40].
Regarding kidney transplantation, good results have been obtained compared with those in other kidney diseases; however, in 2-5% of cases, anti-GBM antibody is produced, leading to rapid graft loss [41]. There is no curative therapy for hearing impairment; only symptomatic treatment is available, such as the use of a hearing aid.
Some new drugs have entered clinical trials, such as bardoxolone methyl (Phase II/III) and RG-012 (effect on microRNA-21 interference, Phase II). Other therapies that have thus far been reported to show effects by in vivo and in vitro trials are as follows: (1) treatment with stem cells [42] and (2) treatment to control intracellular signal transduction (USAG-1) [43]. Further expected treatments are as follows: (1) nonsense read-through therapy [44] and (2) exon skipping therapy with antisense oligonucleotides or chemical compounds [45,46]. Notably, our group is working to develop exon skipping therapy using antisense oligonucleotide (ASO). ASO will bind to the exonic splicing enhancer region and disturb the targeted exon at the point of splicing; this will lead to exon skipping. As described above, when the skipped exon nucleotide number is a multiple of 3, it change the severe phenotype with nonsense mutation to a milder phenotype with an in-frame deletion. We have already obtained good results in vitro and are proceeding to an in vivo trial with an animal model.

The border between TBM and AS
An issue of particular interest among experts in the field of AS is the differential diagnosis between AS and TBM. AS has generally been defined as progressive nephritis, sensorineural hearing loss, and specific eye abnormalities; TBM is generally defined as hematuria with/without mild proteinuria, with diffuse thinning of GBM that does not result in renal failure. There remains no clear delineation between the two diseases and leading clinical nephrologists are quite confused. Most heterozygous carriers of variants in COL4A3 or COL4A4 or some females with heterozygous mutations in COL4A5 show only hematuria without proteinuria; those cases are not diagnosed with AS, according to its current definition. It is well-known that most cases with benign familial hematuria possess heterozygous mutations in either COL4A3 or COL4A4, and most will not develop ESRD [47]. It is also well-known that, in cases with only hematuria without proteinuria, ESRD will never develop 1 3 [9]. In EM findings, these cases typically show only TBM without lamellation of the lamina densa [7,14]. In contrast, cases with ADAS begin to show proteinuria at the median age of 17 years; for females with XLAS, the equivalent age is 7 years. These findings indicate that there is no clear distinction between these two diseases, which presents difficulties for physicians when attempting to make a differential diagnosis. We have experienced some cases exhibiting only hematuria at a young age, which were diagnosed with TBM; these patients then missed annual check-ups and developed ESRD at approximately 60 years of age (personal experience). ACEI can delay the age of developing ESRD, even in male cases of XLAS with the most severe phenotype. This indicates that ACEI can remarkably delay the age of ESRD development for cases with much milder phenotypes of ADAS. As experts in this field, we are responsible for ensuring that, in AS cases, the opportunity to begin treatment by ACEI is not missed. Diagnosis of TBM tends to underestimate the risk of progressive kidney disease. Therefore, some experts have recently asserted that patients with hematuria, TBM, and heterozygous mutations in COL4A3 or COL4A4 should be classified as cases of autosomal AS. In addition, all females with XLAS should also be classified as cases of XLAS. These changes of definition of AS may eliminate TBM nephropathy as a diagnostic entity [48].

Conclusion
Recent advances in genetic analysis have provided us with a deeper understanding of the pathophysiological mechanisms of inherited kidney diseases. Based on precise genetic information, we came to understand the mechanisms of onset of milder phenotypes in AS. As experts in this field, we must devote our energies to creating a system that allows patients with AS to be diagnosed at an early stage and to start appropriate treatment. For this reason, we need to change the definition of AS to include all cases with nephropathy caused by COL4A3, COL4A4, or COL4A5 gene variants and eliminate the disease entity of TBM; this will allow patients to avoid missing the opportunity to start appropriate treatment as early as possible.