The familial hematuria of glomerular origin is usually due to a group of genetically and phenotypically heterogeneous disorders, such as benign familial hematuria (BFH), Alport syndrome (AS), glomerulopathy with fibronectin deposits (GFND), and C3 glomerulonephritis (C3GN) (Deltas et al., 2013). The common causes of familial hematuria among children and young adults are BFH and AS (Kashtan, 2009; Gale, 2013). BFH (OMIM 141200) is usually a non-progressive autosomal dominant renal disorder with a penetrance of 70% (Blumenthal et al., 1988; Lemmink et al., 1996; Savige et al., 2013), which is mainly characterized by persistent or recurrent glomerular hematuria, diffuse thinning of the glomerular basement membrane (GBM), and normal renal function (Slajpah et al., 2007). BFH is usually referred to as thin basement membrane nephropathy (TBMN) (Marcocci et al., 2009; Deltas et al., 2013; Gale, 2013), but TBMN is no longer considered benign as some can develop into end-stage renal disease (ESRD) in adults (Dische et al., 1985; Naylor et al., 2021). The frequency of this condition is estimated to be at least 1.0% in the worldwide population (Buzza et al., 2001; Chan and Gale, 2015), 5.2% in a population with microscopical examination and morphometry of GBM (Dische et al., 1990), and 15.5%–26% in children with hematuria and renal biopsy (Piqueras et al., 1998; Roth et al., 2001). It seems to be more prevalent in females than in males (Buzza et al., 2001; Savige et al., 2003). BFH is often attributable to heterozygous variants in the collagen type IV alpha 3 chain gene (COL4A3) or the collagen type IV alpha 4 chain gene (COL4A4), arranging head to head on chromosome 2 (Boye et al., 1998; Tazón Vega et al., 2003). AS most often manifests hematuria, proteinuria, and progressive renal failure, sometimes associated with sensorineural hearing loss and ocular abnormalities (Hudson et al., 2003), and the renal pathology is characterized by irregular thickening, thinning, and lamellation of GBM (Savige et al., 2013; Kamiyoshi et al., 2016). The COL4A3/COL4A4 heterozygous variants were also reported to cause autosomal dominant Alport syndrome (ADAS, OMIM 104200) (Boye et al., 1998), while homozygous or compound heterozygous variants were responsible for the autosomal recessive Alport syndrome (ARAS) (Longo et al., 2002; Voskarides et al., 2007). The heterozygous carriers of the ARAS family can present manifestations like BFH (Heidet et al., 2001; Savige et al., 2003). X-linked AS (XLAS), the most common form, accounts for about 85% of AS patients, which is ascribed to variants in the collagen type IV alpha 5 chain gene (COL4A5) (Slajpah et al., 2007; Nabais Sá et al., 2015). Compared with ARAS and XLAS, ADAS is relatively milder and slowly progressive with rare extrarenal manifestations (Kamiyoshi et al., 2016). Given the heterotrimeric association of type IV collagen α3/α4/α5 chain in the GBM and the overlapping clinical symptoms, a spectrum ranging from totally asymptomatic or isolated microscopic hematuria to proteinuria, up to chronic renal failure (CRF) and ESRD (Nabais Sá et al., 2015; Furlano et al., 2021; Kashtan, 2021), it is rational to consider BFH and AS as an entity and to term them collectively as collagen IV-related nephropathies (Tazón Vega et al., 2003; Pieri et al., 2014). BFH is considered to be the mildest end in the continuous spectrum, and the most severe end is represented by ARAS (Lemmink et al., 1996; Nieuwhof et al., 1997; Xu et al., 2016). However, BFH can resemble early AS, especially ADAS, in initial clinical presentation and the electron-microscopic features of the GBM, rendering the exact diagnosis notoriously difficult (Badenas et al., 2002; Gross et al., 2003; Voskarides et al., 2007). A complete investigation of family history, careful clinical evaluation, long-term specialist follow-up, and available renal biopsy analysis are vital in avoiding misdiagnosis or missed diagnosis (Kashtan, 2021). This study aimed to identify the genetic etiologies of familial hematuria in two unrelated Han-Chinese families and to elucidate possible morbigenous mechanisms.

The COL4A4 gene (OMIM 120131), located at 2q36.3, contains 48 exons encoding the α4 chain of type IV collagen (Boye et al., 1998). Each chain contains an N-terminal 7S domain, a long, central, triple-helical collagenous domain of Gly-Xaa-Yaa repeats frequently interrupted by short non-collagenous regions, and a C-terminal non-collagenous (NC1) domain (Figure 4) (Boutaud et al., 2000; LeBleu et al., 2010). The six highly homologous chains of type IV collagen, namely α1(IV) to α6(IV), were encoded by six genes (COL4A1 to COL4A6) (Hudson et al., 2003; Cicuéndez et al., 2021). In mammals, the existence of six distinct α(IV) chains could only assemble into three different trimers, namely α1α1α2(IV), α3α4α5(IV), and α5α5α6(IV) (LeBleu et al., 2010). During fetal development, collagen α1α1α2(IV) is ubiquitously distributed in all basement membranes (BMs) (Sand et al., 2013), while in the mature GBM and BMs of cochlea and eye, α3α4α5(IV) is the major component (Hudson et al., 2003; Suleiman et al., 2013; Savige et al., 2015). The pathogenic variants in any one of the three genes (COL4A3, COL4A4, and COL4A5) may interfere with the developmental switch, resulting in the persistent expression or compensatory increase of fewer cross-linked and more protease-susceptible α1α1α2(IV), which leads to BFH or AS (Suleiman et al., 2013; Clark et al., 2016; Xu et al., 2016). Currently, more than 400 COL4A4 gene variants have been recorded in the Human Gene Mutation Database (HGMD), of which at least 29 and 65 variants are associated with BFH and ADAS, respectively, and three variants can cause both disorders. These different variants are widely distributed in the COL4A4 gene without hot spots described, and the most frequent variants are glycine substitutions (46/97) (Figure 4). The frequency of nonsense, splicing, and frameshift variants in ADAS seems to be higher than that in BFH (44.12% vs. 37.50%), consistent with that these variants usually cause more severe phenotypes (Shang et al., 2019).

In this study, two variants in the COL4A4 gene were identified in two unrelated Chinese pedigrees with familial hematuria. The heterozygous c.595-1G>A variant is completely co-segregated with the disorder phenotype within pedigree 1. The heterozygous c.1715G>C transversion detected in the proband of pedigree 2 has been previously reported in compound heterozygosity with variant c.-23T>G in a Chinese AS male patient, and both his sons with heterozygous c.-23T>G variant only presented isolated hematuria (Liu et al., 2017), supporting that the identified c.1715G>C variant may be pathogenic and seems to exert a pathogenic effect to result in a more severe renal phenotype. The absence of these two variants in the asymptomatic family members, the prediction of its deleterious effect by in silico programs, cDNA analysis, protein sequence alignment, and ACMG criteria evaluation suggested that they were likely disease-causing variants in these two pedigrees.

The splicing variant c.595-1G>A may induce NMD of the altered COL4A4 mRNA or produce a truncated loss-of-function α4(IV) chain, consistent with similar studies (Schwarze et al., 2001; Davidson et al., 2007; Korstanje et al., 2014; Hashikami et al., 2018). The missense variant c.1715G>C (p.Gly572Ala) occurred in the first position of critical Gly-Xaa-Yaa repeats in the collagenous domain. The highly conserved glycine residue is the smallest amino acid that can fit tightly into the center of the triple helix structure of the α(IV) chain and is crucial to the helix formation (Deltas et al., 2013; Murray et al., 2014). Hence, both variants in the COL4A4 gene may disrupt the normal synthesis of the α4(IV) chain and cause defective synthesis, assembly, deposition, or function of the α3α4α5(IV) network (Korstanje et al., 2014; Kashtan et al., 2018), interfering natural development process and leading to the abnormal GBMs (Clark et al., 2016; Naylor et al., 2021). Additionally, the poor association of abnormal α4(IV) with the α3(IV) and α5(IV) chain may leave more misfolded protein in the endoplasmic reticulum (ER), disrupt podocyte’s special secretory capacity, and lead to GBM defects (Pieri et al., 2014). These defective GBMs may partially allow the escape of erythrocytes into Bowman’s space, thereby triggering the onset of hematuria (Yoshikawa et al., 1984).

In pedigree 1, isolated microscopic hematuria, normal renal function, and absence of other signs related to AS in the proband, his mother, and grandfather, are more suggestive of a heterozygous COL4A4-related BFH than AS. In pedigree 2, given that the proband presented microscopic hematuria, mildly elevated serum creatinine level, and irregular thickening and thinning of the GBM, without signs of sensorineural hearing loss or ocular lesions, a diagnosis of early ADAS cannot be ruled out. The IgA deposits within the mesangium corresponding to IgA nephropathy could be due to the predisposition of mesangial IgA deposition in defective GBMs caused by COL4A4 variant (Savige and Harraka, 2021). Usually, from clinical phenotype and renal biopsy, the coexistence of AS and IgA nephropathy is attributed to two independent causes. However, some variants in the COL4A3, COL4A4, or COL4A5 gene seem to increase the susceptibility to IgA nephropathy, which is supported by IgA deposits occasionally observed in the AS patients and the COL4A3, COL4A4, or COL4A5 gene variants reported in a minority of IgA nephropathy patients (Kamiyoshi et al., 2016; Li et al., 2020; Cambier et al., 2021). Based on the identified responsible variant, COL4A4-associated nephropathies, a sub-classification of the conditions in the two pedigrees, is strongly recommended. In this way, these similar or overlapping clinical manifestations could be attributed to a single disease spectrum rather than several distinct disease statuses. It would, therefore, be beneficial to risk assessment, disease management, and genetic counseling of diseased individuals.

Taken together, a novel c.595-1G>A variant and a known c.1715G>C (p.Gly572Ala) variant of the COL4A4 gene were identified in two Chinese families with familial hematuria, respectively. WES has been proven to be a powerful tool for uncovering the genetic etiologies of heterogeneous disorders and identifying at-risk individuals, which will assist in early clinical diagnosis and accurate disease classification. Currently, there is a lack of evidence-based treatment for persistent microscopic hematuria (Xu et al., 2016), but it is of paramount importance to regularly monitor proteinuria development, renal function, and blood pressure every 1-2 years (Savige et al., 2003). Our findings broadened the variant spectrum of the COL4A4 gene and may assist in reproductive risk counseling of these two families and improved medical care.