P-Amyloid-Induced Cerebrovascular Endothelial Dysfunction

TOM THOMAS," CHRIS McLENDON, E. TRUITT SUTTON, AND GEORGE THOMAS

Departments of Psychiatry and Physiology, College of Medicine, University of South Florida, Tampa, Florida 33613, USA

Deposits of amyloid P-peptide (AP) in senile plaques and cerebral blood vessels is a prominent neuropathologic feature of Alzheimer's disease (AD), regardless of genetic predisposition. The cellular origin of cerebral deposits of AP or its precise role in the neurodegenerative process has not been established. Amyloid angiopathy with vascular degenerative changes in the smooth muscle, endothelium, and base- ment membrane could be a precipitating factor for the inflammatory response usually observed in the AD brain.' Recently we demonstrated a novel action of /3- amyloid on peripheral blood vessels, vasoactivity, and endothelial damage through the production of reactive oxygen species? In this report we examine the in vitro effect of AP on cerebral blood vessels in terms of the function and morphology of the endothelium.

METHODS

Freshly isolated bovine mid-cerebral arteries were sectioned into ring segments of 3 mm in length. The tissue was then mounted on hooks attached to force- displacement transducers and equilibrated under optimum tension of 2.0-2.5 g in a 5-ml tissue bath containing oxygenated Kreb's buffer at 37" C. The viability of the blood vessel was established by contracting the tissue using the vasoconstrictor serotonin (5-HT) and by subsequently relaxating it with the endothelium-dependent vasodilator bradykinin. The endothelium was removed mechanically from some blood vessels and the absence of endothelium was confirmed by diminished relax- ation to bradykinin. The contraction and relaxation were repeated after a 15-minute incubation with P-amyloid MI or SOD [150 unitdml]. Segments of the cerebral arteries were also processed for electron microscopy after incubation at 37" C for 30 minutes under various conditions. All experiments were repeated four or more times with blood vessels from different animals. Bradykinin, serotonin and other chemicals were obtained from Sigma Chemical Co. (St. Louis, MO). SOD was purchased from Alexis Biochemicals (San Diego, CA). The source of the P-amyloid peptide was Research Biochemicals International (Boston, MA).

Address for correspondence: Tom Thomas, M.D., PhD., Woodlands Medical and Research Center, 3150 Tampa Road, Oldsmar, Florida 34677. Phone, 813/786-5587; fax, 813/785-3254.

447

448 ANNALS NEW YORK ACADEMY OF SCIENCES

FIGURE la. Enhancement of vasoconstriction by P-amyloid. Bur graph demonstrates con- traction with 5-HT [lO-'O MI under control conditions and after the addition of P-amyloid

MI. Pretreating the tisssue with the oxygen radical scavenging enzyme superoxide dismu- tase (SOD) [150 units/ml], approximately 30 seconds prior to the addition of P-amyloid, nearly completely abolishes the enhanced vasoconstrictor rcsponse to 5-HT. Similarly, vessels in which the endothelium had been mechanically removed demonstrated only minimal in- creases to subsequent 5-HT-induced contractions after incubation with P-amyloid (1-40)

MI. Values are expressed as percentage of maximal contraction to 5-HT in the absence of 0-amyloid protein. Values represent the mean 2 SEM of four or more experiments. Points marked with an asterisk (*) are significantly different from control values ( p < 0.05).

6.0

A 5 4.0

8 'P fi 20

0.0

CONTROL

4- 5.0 10.0 15.0 28.0 25.0

Time (min)

FIGURE 1B. Representative trace showing diminished relaxation after Ap. Vessels were preconstricted submaximally with the vasoconstrictor U46619 [5 X M] and relaxed with increasing doses of bradykinin [5 X M-lO-'M]. The arrows indicate the points of addition of bradykinin under control conditions.

THOMAS et al.: CEREBROVASCULAR ENDOTHELIAL DYSFUNCTION 449

450 ANNALS NEW YORK ACADEMY OF SCIENCES

RESULTS

FIGURE l a illustrates the effect of P-amyloid on contraction of mid-cerebral artery induced by 5-HT. AP significantly enhanced the contraction. Pretreatment with SOD blocked the effect of AD. In the endothelium-free blood vessel ring AP did not potentiate the contraction. The endothelium-dependent vasodilator bradykinin produced relaxation of the cerebral artery in a dose-dependent manner (FIG. lb). Treatment with A@ significantly reduced the relaxation response. The inhibition of relaxation by AP was significantly blocked by pretreatment with SOD. Electron microscopic examination of cerebral arteries treated with AD [lo-‘ MI showed significant damage to the endothelial layer which was prevented by pretreat- ment with SOD (FIG. 2) .

DISCUSSION

Our results show that amyloid P-peptide caused a loss of endothelial function as evidenced by increased vasoconstriction and decreased endothelium-dependent vasodilation. The action of AD was dependent on the presence of endothelium. The apparent loss of endothelial function was verified by electron micrographs of AP-treated cerebral vessels which show destruction of endothelial cells. The oxygen radical scavenging enzyme superoxide dismutase (SOD) prevented the deleterious effects of A@ on cerebral arteries. This indicates that the cerebrovascular effects of AP are mediated by superoxide radicals. Endothelial cells regulate many critical functions including inflammation and blood-brain barrier. Impaired endothelial function may be associated with a variety of vascular disorders associated with aging such as hypertension, cerebral ischemia, vasospasm or stroke? The rapid vascular effects of AP reported here are produced by low levels of soluble peptide and are distinct from the reported neurotoxic effects produced by aggregated forms of AP. The A/3-a used in these studies is the peptide associated with cerebrovascular deposits and is the major form of circulating amyloid.4 The findings of our in uitro study showing endothelial disruption by AP are similar to the swelling and degeneration that are distinctive hallmarks of Alzheimer’s disease and these could lead to neurodegeneration and impaired cognitive function. The significance of interaction of P-amyloid with the endothelium is further illustrated by the recent demonstration of AP binding cell surface receptors on vascular endothelial cells in addition to neurons and microglia.’ We propose that amyloid-mediated vascular damage could be an early event relative to the development of AD, where neurons become vulnerable to free-radical damage. The cascade of degenerative changes would become progressive if antioxidant defense mechanisms are compromised because of aging or environmental factors.

SUMMARY

Cerebrovascular effects of /3-amyloid were investigated using bovine mid-cere- bra1 arteries. P-amyloid-induced endothelial damage was evidenced by increased vasoconstriction, diminished vasodilation and was evident on electron microscopy. The endothelial dysfunction was mediated by reactive oxygen radicals. Vascular damage by P-amyloid may be an early event in the development of the pathology of Alzheimer’s disease.

THOMAS et al.: CEREBROVASCULAR ENDOTHELIAL DYSFUNCTION 451

REFERENCES

1. KALARIA, R. N. 1993. Cerebral microvasculature and immunological factors in Alzheimer’s

2. THOMAS, T. et al. 1996. P-amyloid-mediated vasoactivity and vascular endothelial damage.

3. MAYHAN, W. G. et al. 1990. Effects of aging on response of cerebra1 arterioles. Am. J.

4. YANKNER, B. A. 1996. New clues to Alzheimer’s disease: Unraveling the roles of amyloid

5. YAN, S. D. et al. 1996. Rage and amyloid-0 peptide neurotoxicity in Alzheimer’s disease.

Disease Clin. Neurosci. 1: 204-211.

Nature 380: 168-171.

Physiol. 258: H1138-H1143.

and tau. Nature Med. 2: 850-852.

Nature 382: 685-691.