For years there has been a somewhat murky connection between Alzheimer disease and cerebrovascular disease, and some recent findings thicken the plot even further. Researchers report that certain genetic variations in the gene for SORLA—a lipoprotein receptor implicated in Aβ production—specifically associate either with cerebrovascular disease or with neurodegeneration. The findings suggest that SORLA may influence AD by way of dual mechanisms. Another study suggests that besides the well-known damage vascular Aβ deposits inflict on blood vessels in cerebral amyloid angiopathy (CAA), soluble Aβ itself can also cause cerebrovascular damage. Coupled with other reports that SORLA interacts with the ApoE receptor LRP, and that knocking out SORLA worsens pathology in an AD mouse model, these factors now appear more tightly intertwined with AD and vascular disease than previously thought.
Also known as SORL1 and LR11, SORLA stands for the sesquipedalian “sortilin-related receptor, low-density lipoprotein receptor class A repeat-containing protein.” It has recently moved to center stage as a possible risk factor in common, late-onset forms of Alzheimer’s. Levels of the protein are reduced in vulnerable neurons in AD (see Scherzer et al., 2004) and it is implicated in trafficking of amyloid-β precursor protein (APP) and regulation of Aβ production (see Andersen et al., 2005 and ARF related news story). More recently, a large collaborative effort led by, among others, Lindsay Farrer at Boston University, Massachusetts, reported that genetic variations around the SORLA gene associated with AD in a number of different populations (see ARF related news story). SORL1 has since moved up to fifth place in the Top AlzGene Results tally. Now, in the current issue of Archives of Neurology, Farrer and colleagues report that certain SORLA single nucleotide polymorphisms (SNPs) and haplotypes are associated with cerebrovascular disease, as well.
Researchers of the MIRAGE study group led by Farrer correlated 30 SORLA SNPs with magnetic resonance imaging (MRI) measures of general brain atrophy, white matter hyperintensities (WMHs), and overall cerebrovascular damage in Caucasian and African American siblings. (WMHs are bright spots in MRI scans that are believed to reflect vascular damage.) Looking at single and 3-SNP haplotypes, first author Karen Cuenco and colleagues found that in Caucasian families, distal SNPs (SNPs 16-21 as numbered in their original genetic association study; Rogaeva et al., 2007) associated with cerebral and medial-temporal lobe atrophy, while more proximal SNPs (1, 6, and 8-10) associated with WMHs and general cerebrovascular disease (CVR). CVR is a measure that reflects a combination of WMH and infarcts. Haplotype analysis showed a similar dichotomy between SNPs located on the proximal/distal ends of the gene. The most frequent haplotypes for SNPs 6-8 and 7-9 associated with increased CVR and WMH, while one haplotype spanning SNPs 8-10 associated with a decrease in both. Haplotypes spanning SNPs 21-29, on the other hand, associated with cerebral atrophy and medial-temporal lobe atrophy. “There appear to be two mechanisms by which SORL1 may act in terms of its pathogenesis in AD—through a neurodegenerative process and through a vascular process,” said Farrer in an interview with ARF, though he noted that the nature of either mechanism is still unclear.
Farrer added that the data help explain a duality that appeared in some of the original SORL1 genetic association studies, namely that there are two distinct regions in the SORL1 gene that associate with AD but do so in different populations. Some groups of people have links with variants in only one region, others with variants in the other region, and some populations—Caucasians, for example—associate with both. “Our paper provides some insight that the causative variants appear to act through slightly different mechanisms,” said Farrer.
The limitations of this study are that the use of siblings may enrich for participants who have more brain atrophy than the general population. In addition, the results may not extrapolate to non-Caucasian groups. Some curious discrepancies with the previous SORLA genetic association data exist, as well. For example, the GCG haplotype in SNPs 8-10 was associated with AD (see Rogaeva et al., 2007 and Lee et al., 2007), but in this study this haplotype appeared to be protective.
In summary, Farrer said that this paper is one further piece of evidence linking cerebrovascular disease to AD. “Independent of our work, there appears to be a growing body of evidence from genetics and other disciplines that this is going to be an important pathway to disentangle. This is problematic because, historically, AD was defined on the basis of absence of cerebrovascular risk factors.”
James Lah of Emory University in Atlanta, Georgia, agreed with this view. “It may be hard to make this into a simple story. Clinically we’ve always known that vascular disease or vascular risk factors and AD associate,” he told this reporter. “I don’t think anyone has come up with a good comprehensive hypothesis for that, but the clinical associations are strong.” In the November 26 Journal of Neuroscience, Lah and colleagues reported that Aβ pathology worsens in mice that lack SorLA, but perhaps not in the way that would have been predicted. First author Sara Dodson and colleagues crossed LR11-negative mice with double transgenic PS1/APP animals (PS1δE9/APPswe). “We expected that reduced LR11 would accelerate amyloid pathology and that did occur,” said Lah. But two characteristics of these crosses were surprising. First, the researchers found that rather than increasing the maximal amyloid load seen in the brain, lack of LR11 hastened its accumulation. Second, the researchers found that plaque pathology was rampant in the cerebellum, a part of the brain that is usually spared in AD. “We were not expecting that at all and I’m not sure what that is telling us,” said Lah. But he noted that in AD, expression of LR11 is relatively preserved in the cerebellum, suggesting that it, or some related factor, may be protective there. “Whatever the protective factors in the cerebellum that discourage amyloid deposition, when you weaken those by eliminating LR11, then you start seeing quite an impressive amyloid load accumulating there,” he said.
In this study the Emory scientists did not look for cerebrovascular effects of LR11 depletion, but Lah said that they will. “In light of the new genetic associations, I would look in younger animals for abnormal vascular deposition,” he said. “I don’t recall seeing particularly heavy vascular amyloid, but these mice have a lot of vascular amyloid to being with,” he added. With regard to white matter hyperintensities, Lah believes that AD mice model these lesions poorly. “We don’t see much amyloid deposition in the deep penetrating vessels, where the white matter hyperintensities are. We usually see microhemorrhage and CAA in a classical hemispheric distribution, which is very different from the distribution of the hyperintensities,” he said.
While the cerebrovascular effects in AD mouse models have generally been ascribed to CAA, researchers led by Gregory Zipfel at Washington University, St. Louis, Missouri, challenge this conventional wisdom. In the December 10 Journal of Neuroscience, first author Byung Hee Han and colleagues found that vascular dysfunction can precede the deposition of Aβ in blood vessels. This study did not address SORLA. Instead, the St. Louis team directly examined vasodilation in very young Tg2576 mice with no visible CAA. The researchers found that even though six-month-old mice had no obvious parenchymal or vascular Aβ deposits in the leptomeningeal vessels (which are most prone to CAA), their vasodilatory responses were half that of age-matched control mice. In 12-15-month-old mice that do have CAA, the vasodilatory responses were down by 85 percent. The blood vessel response in these older animals depended on the extent of their angiopathy, with those vessels having greater than 20 percent CAA showing no vasodilatory response at all. (CAA was determined by measuring the extent of binding of the amyloid ligand Methoxy-XO4 to blood vessel segments.)
These findings suggest that both soluble and insoluble Aβ can impair cerebrovascular function. In support of this idea, the researchers were able to rescue the vasodilatory dysfunction in young (six-month-old) mice by treating them with the γ-secretase inhibitor LY411575. This drug also restored vasodilation in vessels from 12-month-old mice, provided the degree of CAA was less than 20 percent. As the CAA became more severe, LY411575 was less effective.
Exactly how insoluble Aβ damages the vessels is the next question. Han and colleagues found no pathological abnormalities in vascular smooth muscle cells (VSMCs) of six-month-old animals without CAA, and the cells’ morphology, determined by two-photon microscopy, was intact even in vessels with minimal CAA (less than 20 percent). With greater CAA deposition, however, vessels became visibly damaged, and severely so in 15-month-old mice. Likewise, the density of VSMCs was normal in six-month-old Tg2576 mice but went down in animals with severe CAA. The findings suggest that in young animals, subtle effects in the vessels that are not visible with light microscopy compromise their ability to relax and contract.
If it is not exactly clear how soluble Aβ affects blood vessels, then neither is it obvious how SORLA/SORL1/LR11 affects cerebrovascular disease or neurodegeneration. One potential source of influence is via SORLA binding partners. Best known among these is ApoE, which is not only a SORLA ligand but also the strongest risk factor for late-onset AD. In fact, recent data from Brad Hyman’s lab at Massachusetts General Hospital in Charlestown shows that LR11 interacts with low-density lipoprotein receptor-related protein (LRP), another ApoE binding protein. In a paper in press in the journal Neuroscience, first author Robert Spoelgen and colleagues show that, in neurons, the two proteins interact via multiple domains in both their N-terminals and C-terminals. The scientists used immunoprecipitation experiments and fluorescence energy transfer among different LR11 and LRP constructs.
Unexpectedly, the two proteins seem to interact most strongly in perinuclear vesicles. “The heterodimeric ApoE receptors that have been reported are generally viewed as being cell-surface endocytosis receptors, with the various combinations offering different ligand specificities or affinities,” said Lah. “It may be that the model of LR11 interacting with APP as some type of chaperone may also apply with LRP to influence its intracellular trafficking,” he suggested. In fact, both LRP and LR11/SORLA both interact with APP. To make matters even more incestuous, both are processed by γ-secretase. This opens up the possibility that, like Notch and perhaps APP, the intracellular domains of these lipoprotein receptors have biological activity that could be functionally important. “The promiscuity of these receptors is beginning to get a little annoying,” quipped Lah.
Whatever the precise actions of LR11/SORLA will prove to be in future research, it seems clear from these latest studies that it plays important roles in both cerebrovascular function and in APP processing, making it a potential nexus for pathogenic processes that influence AD.—Tom Fagan
- Scherzer CR, Offe K, Gearing M, Rees HD, Fang G, Heilman CJ, Schaller C, Bujo H, Levey AI, Lah JJ. Loss of apolipoprotein E receptor LR11 in Alzheimer disease. Arch Neurol. 2004 Aug;61(8):1200-5. PubMed.
- Andersen OM, Reiche J, Schmidt V, Gotthardt M, Spoelgen R, Behlke J, von Arnim CA, Breiderhoff T, Jansen P, Wu X, Bales KR, Cappai R, Masters CL, Gliemann J, Mufson EJ, Hyman BT, Paul SM, Nykjaer A, Willnow TE. Neuronal sorting protein-related receptor sorLA/LR11 regulates processing of the amyloid precursor protein. Proc Natl Acad Sci U S A. 2005 Sep 20;102(38):13461-6. PubMed.
- Rogaeva E, Meng Y, Lee JH, Gu Y, Kawarai T, Zou F, Katayama T, Baldwin CT, Cheng R, Hasegawa H, Chen F, Shibata N, Lunetta KL, Pardossi-Piquard R, Bohm C, Wakutani Y, Cupples LA, Cuenco KT, Green RC, Pinessi L, Rainero I, Sorbi S, Bruni A, Duara R, Friedland RP, Inzelberg R, Hampe W, Bujo H, Song YQ, Andersen OM, Willnow TE, Graff-Radford N, Petersen RC, Dickson D, Der SD, Fraser PE, Schmitt-Ulms G, Younkin S, Mayeux R, Farrer LA, St George-Hyslop P. The neuronal sortilin-related receptor SORL1 is genetically associated with Alzheimer disease. Nat Genet. 2007 Feb;39(2):168-77. PubMed.
- Lee JH, Cheng R, Honig LS, Vonsattel JP, Clark L, Mayeux R. Association between genetic variants in SORL1 and autopsy-confirmed Alzheimer disease. Neurology. 2008 Mar 11;70(11):887-9. PubMed.
- Spoelgen R, Adams KW, Koker M, Thomas AV, Andersen OM, Hallett PJ, Bercury KK, Joyner DF, Deng M, Stoothoff WH, Strickland DK, Willnow TE, Hyman BT. Interaction of the apolipoprotein E receptors low density lipoprotein receptor-related protein and sorLA/LR11. Neuroscience. 2009 Feb 18;158(4):1460-8. PubMed.
- T Cuenco K, Lunetta KL, Baldwin CT, McKee AC, Guo J, Cupples LA, Green RC, St George-Hyslop PH, Chui H, Decarli C, Farrer LA, . Association of distinct variants in SORL1 with cerebrovascular and neurodegenerative changes related to Alzheimer disease. Arch Neurol. 2008 Dec;65(12):1640-8. PubMed.
- Han BH, Zhou ML, Abousaleh F, Brendza RP, Dietrich HH, Koenigsknecht-Talboo J, Cirrito JR, Milner E, Holtzman DM, Zipfel GJ. Cerebrovascular dysfunction in amyloid precursor protein transgenic mice: contribution of soluble and insoluble amyloid-beta peptide, partial restoration via gamma-secretase inhibition. J Neurosci. 2008 Dec 10;28(50):13542-50. PubMed.
- Dodson SE, Andersen OM, Karmali V, Fritz JJ, Cheng D, Peng J, Levey AI, Willnow TE, Lah JJ. Loss of LR11/SORLA enhances early pathology in a mouse model of amyloidosis: evidence for a proximal role in Alzheimer's disease. J Neurosci. 2008 Nov 26;28(48):12877-86. PubMed.