Mutations in the COL4A5 Gene
To date, numerous reports have described mutations in the COL4A5 gene, the number already being about 300.14–20,121 The mutations include large (14 percent) and small gene rearrangements such as deletions and insertions (3 percent in-frame, 19 percent causing frameshifts) and splicing mutations (18 percent), as well as single base changes (glycine substitutions 32 percent, other missense mutations 6 percent, and nonsense mutations 8 percent). About 85 percent of the mutations, however, are small. Furthermore, complete loss of the gene has been reported in two cases.35,122,123 A striking finding is that the mutations identified in COL4A5 so far are highly dispersed in this huge gene. The same mutation has been found in only a few cases in two or more unrelated kindreds. This sharply contrasts with findings in other diseases, such as cystic fibrosis where approximately 70 percent of the patients have the same mutation.124 However, the dispersion of mutations in COL4A5 resembles the situation reported for osteogenesis imperfecta, a brittle bone disease that involves a variety of mutations in type I collagen genes125 (see Chap. 205).
Different Effects of COl4A5 Mutations to the Type IV Collagen.
The mutations present in the COL4A5 gene in Alport syndrome can usually explain the structural and pathophysiological changes in the GBM. The mutations can result in changes that affect many of the steps of type IV collagen synthesis and assembly, such as: (a) the synthesis of the primary transcript; (b) posttranscriptional modification of the transcript such as splicing; (c) translation of the mRNA; (d) posttranslational modifications of nascent α-chains; (e) assembly and folding of three α-chains into a triple-helical molecule; (f) the stability of the helix structure; and (g) formation of cross-links between individual triple-helical molecules. Large gene rearrangements can result in total loss of expression of the gene and small mutations in regulatory elements can have the same effects. Large and small mutations can also affect splicing of the primary transcript. Abnormal splicing and skipping of exons can cause two types of changes. First, if the exon skipping is in-frame, the result is a shortened polypeptide so that it lacks a segment, second, if the exon skipping occurs out-of-frame, the result is a protein partially containing a nonsense amino acid sequence.
Effects of COl4A5 Mutations to the Type IV Collagen Network.
Mutations leading to absence of the α5(IV) chain or extensive alterations in protein size would certainly lead to abnormal structure of any type IV collagen network where the α5(IV) chain is required as a normal component. Complete loss or presence of an abnormal α5(IV) chain may explain the clinical manifestations observed in Alport syndrome. In the kidney, the α5(IV) chain is normally present only in the GBM as shown by immunostaining techniques, with a similar GBM staining pattern observed for the α3(IV) and α4(IV) chains. Although it is not known in what molecular chain combinations the α5(IV) chain is present in the GBM, it was recently shown by Gunwar et al.97 that α3:α4:α5 network is formed in the GBM. Consequently, the absence of a normal α5 chain may weaken the structural network of the GBM in one way or another, leading to leakage of large proteins or even large blood cells into the urinary space. It is important to note, however, that heterozygous females usually exhibit only mild phenotypes and they seldom develop end stage renal disease. This suggests that the synthesis of low amounts of a normal α5(IV) chain is sufficient to maintain normal GBM function, even though the abnormal chain is present. This fact is particularly important when considering future possibilities for gene therapy through the supplementation of expression of a normal transgene in addition to that of the mutated gene.
Single base mutations in exons leading to amino acid changes in the α5(IV) chain can be expected to cause the protein malfunction in most cases because such changes can affect proteins in a number of ways. The question then is how can we know whether an amino acid substitution actually causes disease instead of being silent. This is an important question, especially from a clinical point of view when making decision about the termination of pregnancy after a mutation in the fetus has been identified. Actually, this is also a very difficult question to answer. One relatively definite way of proving that a mutation is causative for the disease is to generate transgenic animals with the same mutation and observe whether it results in Alport syndrome-like phenotype. However, it is possible that a mutation causing disease in man does not cause the same disease in mice. One, therefore, has to rely on strong “circumstantial” evidence when predicting the potential pathogenic effect of mutations. Such evidence can be linkage or segregation of the mutations with the phenotype in an Alport kindred, knowledge about the role of certain amino acids for the function of the protein, or conservation of the amino acids during evolution. It is assumed that amino acid residues conserved in a protein during evolution between widely distant species have such functional importance that their substitution cannot be tolerated. If mutations occur de novo concurrent with the appearance of disease, they are likely to be causative.
A large number of mutations converting a glycine residue in the Gly-X-Y repeat containing collagenous domain to another amino acid have been published. These mutations can, without any doubt, affect the stability of the triple helix of the molecule because, as mentioned above, glycine is the only amino acid small enough to suit into the center of the triple helix and an uninterrupted Gly-X-Y repeat sequence is essential for the maintenance of the triple-helical conformation. The result of glycine substitution is, therefore, that the helix is destabilized, creating kinks in the molecule that are not tolerated. Although this has not been demonstrated at the protein level for type IV collagen, a number of similar glycine mutations, either inhibiting helix formation or causing kinks in the molecule have been described in type I collagen in osteogenesis imperfecta.125
COl4A5 Mutations Cause Different Phenotypes
Alport syndrome is a heterogeneous disease with onset of end-stage renal disease at varying ages and with or without manifestations such as hearing loss and deafness, ocular lesions, thrombocytopathy, and so on. Because so many mutations have been identified in the COL4A5 gene, it is of interest from a clinical point of view to correlate them with the phenotypes they produce. However, this analysis is disappointing in the sense that it does not seem to be possible to predict the phenotype based on the type of mutation. For example, the point mutations generated by substitution of glycine residues in the collagenous domain or conserved amino acids in the NC1 domain result in a quite wide spectrum of phenotypes. Most, but not all patients have juvenile onset of end-stage renal disease, but the presence of hearing loss and ocular lesions varies between individuals.
A similar heterogeneous picture is seen for large changes in the gene. Even the entire loss of the gene does not necessarily produce more severe disease than a single glycine substitution in the collagenous domain of the α5(IV) chain. Although disappointing, this does not come as a surprise as the same phenomenon has been observed for defects in the genes for fibrillar collagens causing osteogenesis imperfecta, chondrodysplasia, Stickler syndrome, and Ehlers-Danlos syndrome.125 A clinically important group are the ≈15 percent of posttransplantation Alport patients who develop anti-GBM antibodies. Initial studies on Alport mutations and phenotypes gave hope that, based on the nature of mutation, it might be possible to predict which patients develop anti-GBM nephritis and, thus, who should obtain a transplant and who not. It seemed reasonable that a mutation deleting the highly antigenic NC1 domain of the α5(IV) chain would cause the patient to recognize the corresponding domain in the allograft as a foreign protein and should, therefore, not receive an allograft as the patient would develop anti-GBM nephritis. However, more recent data on transplanted patients with similar kinds of gene defects have shown that this explanation is not totally valid. Based on current, yet still limited knowledge, it is not possible to predict which patients will develop anti-GBM nephritis following kidney transplantation.
Mutations in the COl4A3 and COl4A4 Collagen Genes in Autosomal Alport Syndrome
Although Alport syndrome is primarily considered an X-linked disorder, autosomal forms were also postulated in about 15 percent of the cases based on pedigree data.126 Following identification of defects in the α5(IV) collagen gene in the X-linked disease, it was logical to search for mutations in autosomal disease in the COL4A3 and COL4A4 genes, encoding the α3(IV) and α4(IV) collagen chains, respectively, as these chains are also prominent in the GBM. Such mutations have now been found both in the COL4A3 23–25,126 and COL4A4 23,26 genes in several Alport kindreds. The patients have been shown to be either homozygotes or compound heterozygotes for the mutation, and the parents unaffected carriers indicating autosomal recessive pattern of inheritance. Five different mutations have been reported in the COL4A3 gene in autosomal recessive kindreds, present in either homozygous or heterozygous form. Two of these mutations are nonsense,23,24 two are small deletions in exons,23–25 and one is a splice-site mutation.126 All five mutations result in premature translation stop codons, which would lead to truncated malfunctioning polypeptides. The first two mutations found in COL4A4 by Mochizuki and coworkers23 were a nonsense mutation and a glycine substitution. Recently, Boye et al.26 studied 31 unrelated autosomal recessive Alport syndrome patients and found 10 novel COL4A4 mutations in 8 patients. Five of these mutations create premature stop codons or a shift in the reading frame, and are, therefore, potential null mutations. One mutation probably causes a deletion of two Gly-X-Y repeats in the collagenous region, thereby shortening the protein. Two mutations affect splicing and two are missense mutations, one a glycine substitution in the collagenous region and the other a highly conserved proline to leucine substitution in the C-terminal NC1 domain. A missense mutation, substitution of evolutionarily conserved leucine to proline, was recently described in a large family with rare autosomal dominant form of Alport syndrome.22
Immunohistochemistry of Alport Syndrome Affected Tissues
Immunohistochemical studies demonstrate that the six type IV collagen α chains differ quite remarkably in their tissue distribution (Table 214-2). This, in turn, suggests that the different isoforms serve different functions in vivo. The α1(IV) and α2(IV) chains that form the most common α12α2 isoform are present in practically all basement membranes where they are frequently the only chains present. This emphasizes a major general role of this isoform in basement membranes. The other minor chains have a highly restricted distribution.75
Table 214-2: Distribution of the Six Type IV Collagen Chains in Normal Tissues and in Tissues of Alport Patients |Favorite Table|Download (.pdf) Table 214-2: Distribution of the Six Type IV Collagen Chains in Normal Tissues and in Tissues of Alport Patients
| ||Normal/X-linked AS/Recessive AS || |
| || || |
|Tissue ||α1(IV) ||α2(IV) ||α3(IV) ||α4(IV) ||α5(IV) ||α6(IV) ||References |
|Kidney || || || || || || ||21,127–132,134 |
|GBM embryonic* ||+ ||+ ||– ||– ||– ||– || |
|GBM postnatal ||(+)/+/+ ||(+)/+/+ ||+/–/– ||+/–/– ||+/–/– ||–/–/– || |
|Bowman's capsule ||+/+/+ ||+/+/+ ||+/–/– ||+/–/– ||+/–/+ ||+/–/? || |
|Distal tubules ||+/+/+ ||+/+/+ ||+/–/– ||+/–/– ||+/–/+ ||+/–/? |
|Collecting ducts ||+ ||+ ||– ||– ||+ ||+ || |
|Mesanglum ||+ ||+ ||– ||– ||– ||– || |
|Eye || || || || || || ||133 |
|Anterior lens capsule ||+/+/? ||+/+/? ||+/–/? ||+/–/? ||+/–/? ||?/?/? || |
|Internal limiting membrane ||+ ||+ ||+ ||+ ||+ ||? || |
|Bruch's membrane ||+ ||+ ||+ ||+ ||+ ||? || |
|Descemet's membrane ||+ ||+ ||+ ||+ ||+ ||? || |
|Cochlea ||+ ||+ ||+ ||+ ||+ ||? ||131 |
|Skin ||+/+/+ ||+/+/+ ||–/–/– ||–/–/– ||+/–/+ ||+/–/? ||128,130,132,134 |
|Esophagus ||+ ||+ ||– ||– ||+ ||+ ||131 |
|Brain ||+ ||+ ||+ ||+ ||– ||– ||131 |
|Lung ||+ ||+ ||+ ||+ ||+ ||+ ||131 |
The α3, α4, and α5 chains of type IV collagen are expressed in the kidney, and their expression is highly enriched in the glomerular basement membrane, but they are also found in a subset of tubular basement membranes, such as Bowman's capsule and the distal tubules.127,128 The α5(IV) chain is, additionally, present in the collecting ducts. Expression of the α6(IV) chain is almost identical to that of α5(IV) throughout the kidney, but it is notably absent from the GBM.127 In many male patients with X-linked Alport syndrome the α3(IV), α4(IV), and α5(IV) chains are absent from the GBM, whereas expression of the α1(IV) and α2(IV) chains is increased.127,129 Expression of the α6(IV) chain is normally seen in Bowman's capsules and tubules but not in such patients; this is also the case with the α3(IV), α4(IV), and α5(IV) chains.128 Women who are heterozygous for the X-linked disease exhibit mosaic expression of the α3(IV), α4(IV), and α5(IV) chains in the GBM, while expression of the α1(IV) and α2(IV) chains is preserved.130
Immunohistochemical analyses of skin tissues from males with X-linked Alport syndrome have shown a complete absence of the α5(IV) and α6(IV) chains normally expressed in the skin. A discontinuous or mosaic staining for these two chains is seen in the epidermal basement membrane (EBM) of female heterozygotes.128,131,132 Cheong et al.133 have studied the expression of type IV collagen α chains in another Alport target tissue, the anterior lens capsule, in X-linked cases, and demonstrated the same features as in the GBM, namely, a lack of the α3(IV) to α5(IV) chains (Table 214-2).
Gubler et al. studied the distribution of different type IV collagen chains in autosomal recessive Alport syndrome.134 Usually, no expression of the α3(IV), α4(IV), or α5(IV) chains can be seen in the GBM, but deviating from the X-linked expression pattern, the α5(IV) chain is normal in Bowman's capsule and distal tubular basement membrane. A similar expression pattern can be seen in both sexes. In addition, the α5(IV) chain shows normal staining in the EBM, while the α3(IV), α4(IV), and α5(IV) chains are absent from the GBM (Table 214-2). This combination cannot be seen in X-linked AS, as it is well known that only the α5(IV) chain of these three is present in the epidermal basement membrane.134 This differential pattern of staining may be a useful method for distinguishing between X-linked and autosomal recessive forms of Alport syndrome.21