The vascular hypothesis of Alzheimer disease suggests that pathology starts with hypoperfusion, or decreased blood flow to the brain. This, in turn, leads to a crisis among glia and neurons, eventually culminating in neurodegeneration and cognitive impairment (see de la Torre, 2004 and ARF Live Discussion). The theory just got a little shot in the arm. In a Nature Medicine paper published online last Sunday, Berislav Zlokovic and colleagues show that expression of MEOX2, a homeobox gene that is upregulated during vascular differentiation, is significantly reduced in late-stage AD and is required for optimal clearance of amyloid-β (Aβ) from the brains of mice.

Zlokovic, from the University of Rochester, New York, together with colleagues there and at other labs in New York and California, found the link between AD and MEOX2 when they carried out microarray analysis of messenger RNA levels in brain tissue samples. Joint first authors Zhenhua Wu, Huang Guo, Nienwen Chow and Jan Sallstrom—the last three from Socratech Research Laboratories in Rochester—measured levels of almost 13,000 genes in brain endothelial cells (BECs) isolated from frontal cortex autopsies. Wu and colleagues studied six patients with late-stage AD, six age-matched controls (mean age 70 years) having no dementia, and five younger (average age 24) subjects. The authors found about 34 genes that were differentially expressed in the AD samples compared to samples from nondemented and younger individuals.

Homing in on these 34 genes, Wu and colleagues found that MEOX2, encoding the homeobox protein GAX, and TGM2, encoding a transglutaminase, were the most dramatically affected. Quantitative PCR analysis on BECs captured by laser microdissection revealed that MEOX2 expression was down almost 20-fold, whereas TGM2 expression was elevated by about the same degree. Other genes that are involved in angiogenesis were also downregulated (NR4A2 and PLEC1), while the proapoptotic, forkhead transcription factor gene MLLT7 was upregulated.

The loss of MEOX2 expression could have important consequences for vascularization in the CNS and for clearance of Aβ. Wu and colleagues found that if they silenced MEOX2 to reduce GAX to about 40 percent of normal in endothelial cells, then capillaries formed in culture were shorter by about 60 percent while forkhead expression increased about twofold. This suggests that loss of MEOX2 in AD patients might be driving endothelial cell apoptosis and loss of capillaries. Furthermore, the authors found that MEOX2+/- heterozygote mice had about a 50 percent reduction in cerebral blood flow at 2-3 months of age, and they were unable to mount an angiogenesis response to hypoxia; normal mice increased total cortical capillary length by about 40 percent after 3 weeks of hypoxia, whereas in MEOX2+/- animals there was no change in the capillary network. Furthermore, these heterozygotes retained about twice as much Aβ40 in the brain as did normal animals and clearance across the blood-brain barrier of radioactive Aβ40 injected into the brain was reduced by 80 percent.

The failure to clear Aβ may not only be related to poor vascularization. Wu and colleagues also found that when MEOX2 is silenced, lipoprotein receptor-related protein 1 (LRP), a major carrier for Aβ, is reduced by about 40 percent in human BECs (though see ARF related news story which questions the link between LRP and Aβ clearance).

All told, the data seem to support the idea that vascular losses could contribute to AD pathology. While six patients is a small sample set, the authors were able to confirm their transcriptional profiling data using samples from a further ten patients in late-stage AD and nine age-matched controls having no AD-like pathology. Also of interest is that the authors found no changes in GAX levels in Tg2576 mice which, although rife with Aβ deposits, show very little neurodegeneration.

The authors write that low expression of MEOX2 in Alzheimer disease could have two major effects: “(i) impaired angiogenesis associated with apoptosis, vessel malformation and regression, ultimately resulting in reductions in brain capillary density and CBF [cerebral blood flow], as seen in Alzheimer disease and (ii) a pathological BBB [blood brain barrier] phenotype with little or no Aβ-clearing capability resulting from low levels of LRP, which may lead to Aβ accumulation as seen in Alzheimer disease models and the disease itself.”—Tom Fagan


  1. Prior to the 1980s, the cause of death of elderly demented individuals was commonly listed as pneumonia. In the 1980s, Alzheimer disease gained increasing attention, in part due to the tireless efforts of pathologists, such as Robert Terry, and neurologists, such as Robert Katzman. Terry and Katzman made the public and the medical community aware that senile dementia of the Alzheimer type was not the normal form of aging, was also not the normal outcome of other diseases, such as vascular diseases (hypertension, multi-infarct dementia, etc.) or diabetes, and should be listed as a principal cause of death, even when the immediate cause of death was a disease, such as pneumonia. In the intervening years, the field has advanced tremendously, and research has identified the roles of Aβ and tau, and codified the amyloid cascade hypothesis.

    Despite the mechanistic clarity provided by the amyloid cascade hypothesis, many questions remain, particularly with respect to late-onset Alzheimer disease. At the same time, increasing evidence indicates that other chronic diseases, such as hypertension, mid-life hypercholesterolemia, and diabetes, increase the incidence of AD and might contribute to the pathophysiology. The current manuscript by Wu and colleagues provides some of the molecular details that might contribute to the vascular-Alzheimer linkage. In an elegant study, they did transcriptional profiling on brain endothelial cells, and identified transcripts whose expression was not altered in normal aging but was altered in the Alzheimer brain. Among the genes is a homeobox gene, termed MEOX2, which acts as a transcription factor and regulates a number of genes that could easily be important for the pathophysiology of AD. MEOX2 regulates factors important for Aβ uptake, such as LRP1, factors important for apoptosis, such as AFX1, and factors important for capillary growth. They follow the amyloid story into genetically engineered mice, and show that mice lacking MOEX2 exhibit reduced flux of Aβ across cerebral endothelial cells. These experiments are very intriguing and provide an interesting mechanism that could link AD with vascular abnormalities. The concept of Aβ flux is also particularly appealing because this idea was highlighted by studies on passive Aβ immunization. On the other hand, MEOX2 is involved in so many different processes that it is likely only a fraction of them are truly relevant to AD. Defining how vascular abnormalities contribute to AD clearly will be a fascinating avenue of research as the field matures and is able to incorporate new mechanistic details into our understanding of late-onset AD.

  2. Wu and colleagues, in a recent Nature Medicine paper, report a potential role for the MEOX2 hemeobox gene, also known as GAX, in Alzheimer disease. The authors cultured brain endothelial cells (BECs) from humans who died with Alzheimer disease, age-matched controls, and young controls. They found upregulation and downregulation of groups of genes in the BECs derived from AD cases. A gene that was strongly downregulated was MEOX2. Due to the previously described role of MEOX2 in angiogenesis, they further explored whether MEOX2 may have a role in brain angiogenesis. They found that in MEOX2 +/- mice, there was a decrease in cerebral blood flow and capillary length, and that vessels from these mice did not respond normally to hypoxia. Since the blood-brain barrier (BBB) is involved in amyloid-β (Aβ) efflux from the brain, they also determined whether there was a deficit in Aβ40 efflux out of the brain. A large decrease in Aβ40 was found that was consistent with a decrease in efflux mediated via BBB transport. The fact that LRP was also decreased in brain capillaries from these mice may be why there was decreased Aβ40 efflux. Overall, these results suggest that a decrease in MEOX2, if present in AD brain capillaries in situ, may be playing a role in decreased capillary density and other important changes in the AD brain that could contribute to cognitive deterioration in AD. In terms of AD pathogenesis, the authors did not observe a decrease in MEOX2 in brain capillaries from an APP transgenic mouse model that develops amyloid deposition. This suggests that the change in MEOX2 in AD may not occur as an early step in AD pathogenesis since Aβ deposition appears to occur very early in the pathogenic cascade that ultimately culminates in the cognitive changes seen in AD. Understanding whether and how certain molecules contribute to AD progression is important, and further studies to sort out whether MEOX2 contributes to AD progression are certainly warranted. It will be interesting to see if there is a change in amyloid-related pathology if MEOX2 +/- mice are bred to APP transgenic animals. It will also be interesting in future studies to know the effect of enhancing MEOX2 function in other AD-type models as well as in models of other neurodegenerative diseases.

  3. Excellent article.

  4. This hypothesis is old and dead. Of course hypoperfusion can increase the risk of AD, but is not shown to be strongly causative, or necessary, in amyloidosis of either the brain or other organs. Reduced blood supply to the carotid artery can, in some cases, evoke a strong immune response or be triggered by monocyte-derived cholesterol, but is not shown definitively to be a key ingredient in AD pathogenesis. Furthermore, reduced blood supply to neurons may, in many cases, have an opposite effect in that where there is lack of immunoglobulins and oxygen, Aβ42 cannot be triggered through the presenilins' control of secretase.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Lack of Lipoprotein Receptor Boosts Brain ApoE, but Not Aβ

Paper Citations

  1. . Is Alzheimer's disease a neurodegenerative or a vascular disorder? Data, dogma, and dialectics. Lancet Neurol. 2004 Mar;3(3):184-90. PubMed.

Other Citations

  1. ARF Live Discussion

Further Reading

No Available Further Reading

Primary Papers

  1. . Role of the MEOX2 homeobox gene in neurovascular dysfunction in Alzheimer disease. Nat Med. 2005 Sep;11(9):959-65. PubMed.