$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
Age-associated diseases are becoming increasingly prevalent in modern society. As medical science improves and life expectancy increases, the population continues to age and the burden of these diseases grows. Effective treatments are necessary to lessen the impact of debilitating conditions but remain elusive in many cases. Only by understanding the causes and pathology of diseases associated with aging can scientists begin to identify potential therapeutic targets and develop strategies for intervention. Common age-related conditions include neurodegenerative disorders such as Parkinson's disease (PD) and age-related macular degeneration (AMD). PD is the most common movement disorder caused by neurodegeneration in humans. Most PD patients show symptoms such as resting tremor, bradykinesia and rigidity after 50 years of age. Early onset has also been observed in approximately 10% of cases.
AMD is the leading cause of blindness in the elderly, progressively damaging photoreceptors and the retinal pigment epithelium (RPE) in the eye. Central vision is impaired but the peripheral vision is generally unaffected. There are two forms of AMD. In the "dry" form, extracellular protein deposits known as drusen form between the RPE and Bruch's membrane (BM), leading to geographic atrophy and blurring of central vision. The more severe "wet" form results from neovascularization from the choroid across BM into the RPE and photoreceptor layers and can lead to hamorrhaging beneath the retina that causes permanent damage to retinal tissue. Genome wide association studies have previously identified loci that are associated with increased susceptibility to this disease and identified two regions of interest: complement factor H (CFH) on chromosome 1 and the 10q26 locus, where the age-related maculopathy susceptibility 2 (ARMS2) and high-temperature requirement factor A1 (HtrA1) genes are located1-5. Combinations of these alleles increase the likelihood of AMD in a dose dependent manner and specific SNPs can be preferentially associated with either the wet or dry forms of AMD3-6.
CFH acts as a negative regulator of the alternative pathway (AP) of the complement system by inhibiting the activation of C3. A single nucleotide polymorphism (SNP) has been linked to increased risk of AMD, causing exchange of tyrosine 402 in exon 9 with histidine due to a T to C substitution1. In AMD it is believed that the AP is misregulated because of a loss of function of CFH but whether the SNP plays a causal role is unclear. One hypothesis is that the positively charged histidine is thought to negate the ability of CFH to bind to interacting proteins C-reactive protein and heparin sulfate1,7. In vitro studies of CFH Y402H provide conflicting results over functional differences between the variants, and in vivo work in Cfh-/- mice expressing humanized CFH is ongoing8. ARMS2 is located upstream of HtrA1 in the primate genome, although the gene is absent in mice. HtrA1 is a serine protease but ARMS2 is poorly characterized. The linkage disequilibrium between SNPs in the AMD-associated locus has made it difficult to determine the contributions to risk of individual mutations of the genes in this region, but recent work has suggested that it is overexpression of HtrA1 rather than ARMS2 that leads to neovascularization and subretinal protein deposits9-11. However, the close proximity of the genes in this locus may allow for interactions that cannot be studied using randomly inserted transgenes.
In addition to AMD, the HtrA family of serine proteases has been associated with many human diseases. All HtrA proteins contain a serine protease domain followed by at least one C-terminal PDZ domain. HtrA1, HtrA3 and HtrA4 share the greatest homology, consisting of a signal peptide, insulin-like growth factor binding domain, a Kazal protease inhibitor domain, the serine protease domain and a PDZ domain. HtrA2 has a different N-terminus, composed of a mitochondrial localization sequence, transmembrane domain and inhibitor of apoptosis binding domain followed by the protease and PDZ domains12-16. Mammalian HtrA1 is regulated by substrate-induced remodeling in the active site of its protease domain17-20, and HtrA2 can also be modulated by interaction between the serine protease and PDZ domains that suppresses protease activity21. Interestingly, the PDZ domain does not appear to confer similar regulation to HtrA316. The HtrA proteases can also be regulated by extrinsic factors: it was recently demonstrated that there exists a regulatory interaction between HtrA1 and protoporphyrins22 and HtrA2 can be regulated by phosphorylation upon activation of the p38 MAP kinase pathway in a PINK1-dependent manner23. The deletion of individual members of the HtrA family in mice has been documented, however mechanistic effects mostly are unclear partly due to lack of visible phenotypes.
HtrA1 plays an important function in protein quality control and its misregulation or mutation has been associated with many different human diseases including arthritis, cancer and an increased risk of AMD3,4,24-32. Loss of HtrA2 function in neural tissues has been associated with PD phenotypes in humans and mice, while its loss from non-neural tissues results in accelerated aging33-37. HtrA3 dysregulation has been associated with diseases including preeclampsia and certain types of cancer38,39. Up-regulation of HtrA4 has been observed in the placentas of preeclampsia patients but knockout mice do not display an overt phenotype40,41. The lack of phenotypes observed in some knockout mice has been postulated to be a result of compensation between the HtrA family member: it is thought that both HtrA4 and HtrA1 interact with the TGF-B family of proteins, allowing for compensation by HtrA1 upon deletion of HtrA441. Similarly, it is thought that since HtrA1 and HtrA3 have a high degree of domain homologies they might have complementary functions42. However, it has been suggested that HtrA proteins may have partially antagonistic roles, competing to regulate common targets43.
To further investigate these risk factors three humanized knock-in mouse lines were generated. In Cfhtm1(CFH*9)jhoh and Cfhtm2(CFH*9)jhoh, exon 9 of the Cfh gene is replaced with exon 9 of the human homologue. Cfhtm1(CFH*9)jhoh encodes the non-disease associated tyrosine residue at position 402, whereas Cfhtm2(CFH*9)jhoh carries the Y402H, risk-associated SNP. In ARMS2tm1jhoh the human ARMS2 sequence was targeted to a region upstream of HtrA1. A loxP-flanked STOP sequence placed upstream of the gene sequence but downstream of the included UbiC promoter was excised by crossing to OzCre mice, which express Cre recombinase under the control of the Rosa26 promoter, as previously described34. In addition to these knock-in lines, conditional knockout alleles for Cfh and HtrA1 (Cfhtm1jhoh and HtrA1tm1jhoh), as well as the other known HtrA family members were generated: HtrA2 (HtrA2tm1jhoh), HtrA3 (HtrA3tm1jhoh) and HtrA4 (HtrA4tm1jhoh). The germ-line knockouts were created by crossing OzCre mice to animals engineered to flank specific exons with loxP sites, such that deletion causes a frame shift and/or deletion of the active domain (Cfh; exon 3, HtrA1; exons 2-3, HtrA2: exons 2-4, HtrA3; exon 3, HtrA4; exons 4-6)34,41. A neural deletion of HtrA2, deleted using Cre recombinase under control of the Nestin promoter (HtrA2flox;Tg(Nes-cre)1Kln/J), has also been described34. Only mice lacking HtrA2, either systemically or in neural tissues, displayed clear phenotypes, presenting with Parkinsonian phenotypes.
Since some of these genes of interest are posited to be localized to mitochondria11,44-47, and deletion of HtrA2 generated Parkinsonian phenotypes, a phenotyping regimen with a mitochondrial and neurological focus is described here and representative results from the tests of interest are provided. In order to examine genetically engineered mice produced to investigate human, age-related disease thoroughly and systematically an initial phenotyping schedule was established, consisting of a series of tests related to the two main diseases of interest: AMD and Parkinson's.