the vasoactivity of aβ peptides

The Vasoactivity of AP Peptides FIONA CRAWFORD," ZHIMING SUO,"

CHUNHONG FANG," ASAD SAWAR," GEORGE SU,' GARY ARENDASH,' AND MIKE MULLAIPb

Departments of "Psychiatry, bNeurology, and 'Biology, Roskamp Laboratories, University of South Florida, 3515 East Fletcher Avenue, Tampa, Florida 33613

ABSTRACT We have demonstrated that freshly solnbilized AP peptides can enhance vasoconstriction by phenylephrine or endothelin of isolated rat aorta. Concentrations of peptide producing these effects (100 nM-1 pM) are much lower than those requiring toxicity to endothelial cells in culture, and effects are immediate, not requiring the prolonged time periods for aggregation necessary in AP cell culture toxicity experiments. Pre-treatment with SOD diminishes the enhancement of vasoconstriction by AP peptides, suggesting that the effects are partly mediated via a decrease in the nitric oxidelsuperoxide ratio. Enhancement of endothelin vasoconstriction is observed with AP1-40 and AP,_,, but not with A&x even at 5 pM, again suggesting the mechanism of AP vasoactivity is distinct from that of AP cytotoxicity. These Observations raise the possibility that AP peptides in contact with the cerebrovascnlature could result in vasoconstriction, hypoperfnsion and oxygen free radical imbalance contributing to the nenrodegeneration of AD.

BACKGROUND

Many data have accumulated that suggest that aggregation of AP peptides is a prerequisite for cytotoxicity in cell culture experiments. This body of knowledge partly supports the idea that in life, aggregation of AP followed by fibril formation and deposition are the steps that precede the neuronal loss resulting in the clinical features of Alzheimer's disease (AD). Alternative views to this schema as the central etiology of AD abound. It is feasible, for instance, that AP peptides have normal functional roles in life, although such roles have not been well established. Physiologic roles would be expected of soluble forms of AP and are therefore unlikely to include complex aggregated forms or fibrillar forms, which are more prone to precipitation. We have examined a potential functional role for freshly solubilized AD peptides in the vasculature. It has been well established (see other papers in this volume) that deposition of AP peptides occurs in the vasculature of AD brains and in the brains of victims of other related conditions such as cerebral amyloid angiopathy. We have shown vasoactive properties of AP peptides which do not require AP deposition and occur when freshly solubilized synthetic AP peptides are applied to living large vessels from (in this report) rats. The concept of AP peptides as vasoactive substances has only been recently supported by experimental evidence' and there are no data as yet connecting these preliminary in vitro observations with a pathophysiologic role in the AD process. Nevertheless, the observations that these peptides, which are widely regarded as central to AD, exhibit these properties suggest that pathophysiologic links of vasoactivity with the disease process might exist and should be sought. However, our data demonstrate actions of AP peptides which are potentially also beneficial. Vasoactivity is a prop- erty of many acute proteins involved in inflammatory and protective roles, and the vasoactive properties of AP may be such an example. Alternatively, the fact that

35

36 ANNALS NEW YORK ACADEMY OF SCIENCES

free radicals, particularly nitric oxide (NO) and superoxide (OF), are intimately associated with vascular tone raises the possibility that other vasoactive substances (such as Ap) might perturb the natural balance between them, resulting in potentially damaging products such as peroxynitrite (ONO;). We should also consider that the action of AP peptides early in life might be wholly beneficial, but that the same action in late life, when the cellular environment becomes more hostile (for instance under more oxidative stress), might be deleterious.

We have observed AP vasoactivity in several experimental paradigms described below which employ a model of isolated rat aorta. The mechanism of the vasoactive properties that we describe is not known, but we have also demonstrated endothelial cell toxicity in cell culture by A P peptides and we contrast the parameters (doses, timing and amyloidogenic nature of the vasoactive fragments) of vasoactivity and cytotoxity and conclude that they are fundamentally different. The implications of this difference for a potential pathophysiologic role of vasoactive AP peptides in AD is discussed. In particular, the observation that API-42 and are both capable of enhancement of vasoconstriction, whereas only the former is toxic to human aortic endothelial cells, is highlighted by the data from early-onset genetic cases (p-APP, PS-1 and PS-2 mutants), who all show increases in serum levels of

which has been suggested as the common feature of their pathogenicity. Our findings might suggest that both and A&,, in solution perform a normal role in enhancing vasoconstriction, but that the former is able to induce vascular cellular damage and may be the key component in the vascular damage seen in AD.

METHODS

Vasoactivity Assay

Vasoconstriction and vasodilation were measured in rat aorta using the system previously described by Thomas et al.’ Normal Sprague-Dawley rats were sacrificed by decapitation, and freshly dissected rat aorta was segmented into rings and sus- pended in Ringer’s buffer using a tensiometer linked to the MacLab system. Follow- ing a 2-hour equilibration period, the “pre-treatment” contraction was carried out. The aortic rings were contracted using a range of doses of phenylephrine (PE) or endothelin-1 (ET-1) and relaxed using either a range of doses of acetylcholine (Ach) as described in Thomas et al. or a single dose (lo-’ M) in the endothelin experiments. After a further 2-hour equilibration, the test compound(s) was added to the buffer surrounding the rings and the contraction repeated at the same doses as before. The dose of each AP peptide was 1 pM, with the exception of (1 p M and 5 pM), and was added 10 minutes prior to the first ET-1 dose. For superoxide dismutase (SOD) treatment we added 150 U/ml30 seconds before the addition of A@. Verapamil was used at 50 pM and was added 2 minutes before the addition of AP.

Endothelium was removed by gently passing stainless steel wire through the lumen of the aortas several times, allowing the wire to gently brush the endothelial surface. Endothelium was removed from the aortas prior to hanging in the tissue bath system.

In the endothelin experiments the mean difference in percentage contraction in the pre- and post-contractions (before and after addition of test compound, respectively) was measured and compared. Percentage contraction was measured by taking the baseline measurement from the contracted measurement and dividing by the baseline measurement.

CRAWFORD el al.: VASOACTIVITY OF PEPTIDES 37

Cell Cultures

A human aortic endothelial cell line (HAEC) was obtained from Clonetics, and grown in endothelial cell growth medium (Clonetics) containing endothelial cell basal medium supplemented with 10 ng/ml human recombinant epidermal growth factor, 1 pg/ml hydrocortisone, 12 pg/ml bovine brain extract, 2% FBS, 50 pglml gentamicin and 50 nglml amphotericin B. A human neuroblastoma cell line (BE(2)- Ml7) was kindly provided by Dr. R. A. Ross and Dr. B. Spengler (Fordham Univer- sity) and grown in DMEM with nonessential amino acids supplemented with 15% FBS (heat-inactivated) and 1 X antibiotic-antimycotic. All these cells were main- tained at 37"C, under an atmosphere containing 5% C02 . For the present experiments, HAEC and BE(2)-MI7 were seeded at densities of 2 X lo4 cells per well in 24-well clusters. The subcultured HAEC was used directly for tests the next day, whereas BE(2)-MI7 was differentiated with retinoic acid and then subjected to tests.

Cell culture media, fetal bovine serum (FBS), and other culture reagents were supplied by Clonetics, Gibco and Sigma. AP1-40 and were supplied by RBI and/or MD Enterprise, and was obtained from Sigma. SOD was obtained from Sigma. Verapamil and pimozide were purchased from RBI.

Experimental Treatment

A@ Toxicity and its Prevention in HAEC

HAEC at a density of 2 X lo4 cell/well with 1 ml fresh EGM medium was exposed to an increasing concentration of (up to 32 pM), Ap1-42 (up to 32 pM), and (up to 40 pM), respectively. For the prevention of Ap toxicity, the cells were pretreated with verapamil (50 pM), pimozide (10 pM), and SOD (250 U/ml) for 3-5 min, respectively, and then exposed to AD. The experiments were terminated 0.5, 1, 2 and 3 hours post treatment by detachment with trypsinl EDTA solution and followed by a toxicity assay with trypan blue exclusion. In parallel experiments, cells were attached to a round Thermanox (Nunc) at the bottom of the well, and were stopped 3 and 24 hours post treatment by fixation with 4% neutral formalin solution, which was followed by hematoxylin-eosin stain to show morphology.

A/3 Toxicity on BE(2)-M17

BE(2)-MI7 of 1 X lo4 per well in a 24-well plate was differentiated with M retinoic acid in 75% ethanol and then exposed to increasing concentrations of

respectively. Trypan blue exclusion was used for the cell viability assay.

AP1-42 and

Measurement of Cell Damage or Death

Observation of Morphological Changes

Morphological changes in cells were monitored throughout the course of the experiment with an inverse phase-contrast microscope. After termination of the experiments, the cells fixed on the Thermanox were subjected to routine hematoxy-


lin-eosin (HE) stain. The morphological characteristics of toxic cells were observed under light microscope and typical results were recorded by photographs.

Cell Viability Assay

For approximate quantification of AP toxicity on HAEC, a modified procedure for trypan blue exclusion was employed. In brief, the detached cells were resus- pended with an appropriate volume of culture media, and 20 p1 of this cell suspension was thoroughly mixed with an equivalent volume of 4% trypan-blue solution and allowed to sit for 2 min after mixing. Approximately 9 pl of the mixture was transferred to a hemocytometer for cell counting. The cell viability was determined by the equation:

number of unstained (living) cells total cells counted (stained + unstained) Cell Viability (%) = x 100%

In the case of severe toxicity, where some of the cells were lysed, the total cell numbers were adjusted by that of controls. In this experiment, counts for each sample were triplicated.

RESULTS

Details of the enhancement of PE constriction by AP have been previously described in Thomas et al.’ In summary, AP addition to rat aorta subsequently constricted with PE shows significantly enhanced vasoconstriction over non-treated aorta. It was also more difficult to relax the AP-enhanced vasoconstricted vessels with acetylcholine. After addition of (1 pM) a sustained contraction occurred within seconds to minutes in some (but not all) preparations, an effect that we have previously shown can be blocked by superoxide dismutase (SOD). These observations were never made in preparations untreated with AD. However, there are clearly a number of uncontrolled and unknown factors that make the intrinsic constriction experimental design difficult to use to examine the characteristics of AP-mediated vasoactivity. We therefore concentrated on developing assays of AP vasoactivity based on the enhancement of constriction by known vasoconstrictors such as PE and endothelins. We tested to see whether the SOD effect in this system required SOD activity or was simply due to binding of AP by SOD. Heat and H202- inactivated SOD had no effect on attenuation of the AP enhancement, nor did the addition of albumin, suggesting it was SOD’S enzymatic activity that attenuated the AP enhancement of vasoactivity rather than nonspecific binding of AD to the SOD protein. To explain the enhancement of vasoconstriction by AP and its prevention by pre-treatment with SOD we have previously suggested that AP alters the balance between interactions of 0, and NO. These two radicals are mutual scavengers. An excess of the former might be expected to scavenge more NO, thus reducing the potential for relaxation. Similarly, a decrease in the latter might be expected to result in vasoconstriction. To test whether AP peptides are able to directly increase the 0;INO ratio by increasing 0, production we metabolically inhibited specific endothelial enzymes known to produce this radical before treatment with AP peptides. Cyclooxygenase and NOS itself both produce O;, but inhibition of these enzymes did not decrease the enhancement of vasoconstriction by peptide.

CRAWFORD el at.: VASOACTWI" OF PEPTIDES 39

In our cell culture experiments with A@ peptides and human aortic endothelial cells (HAEC), significant cell death occurred within 3 hours of application of freshly solubilized AP1-42 (10 pM). 40 pM was required to bring about a similar percentage of cell death. By striking contrast we were unable to bring about significant cell death with even after one week of exposure to (a single dose of) this peptide. Toxicity to HAEC by and A&.35 was time- and dose-dependent (Suo et aL2; FIGS. 1 and 2). Pre-treatment with SOD or verapamil, a calcium channel blocker, significantly inhibited toxicity by

We examined the generality of the enhancement of vasoconstriction by AP peptides by substituting endothelin-1 for PE (Crawford et al., unpublished observations). A comparison of enhancement by AD peptides of vasoconstriction by ET- 1 is shown in FIGURE 4. AS shown, neither dose of differs significantly from controls at any dose of ET-1. 1 pM is significantly lower than AP1-40 at all but the first two doses of ET-1. 5 pM is significantly lower than A& at the first four doses of ET-1. AP1-40 is higher than controls at all but the first dose, and AP1-42 is significantly higher than controls at all but the first dose, but significantly lower than at doses of ET-1 of 2 X lo-' M and 3 X M. FIGURE 5 shows the effects of age of rat aorta on the enhancement of ET-1 constriction by Aortic strips from animals 7-8 months of age show greater enhancement by

or AP2s-.1s to HAEC (FIG. 3).

viability (Sg) ** kO.01 compared to control % , v-

0.00 I

0.50 1.00

I

I

2.00 3.00

Time (hour)

FIGURE 1. Time-dependency of toxicity of HAEC by AD peptides. (After Suo et al., unpublished observations.)

CRAWFORD et al.: VASOACTIVITY OF PEPTIDES 41

C .- Y g Y a


FIGURE 4. Differential vasoactive effects of AP peptides. x axis = dose of ET-1; y axis = mean difference in % contraction1100 z axis = treatment.

AP1-40 than aortic strips from animals 2-3 months of age. The effects of SOD on the enhancement of ET-1 vasoconstriction by AP1-.," are shown in FIGURE 6. There are no significant differences at any dose of ET-1 between AP1-40 with or without SOD. There is no significant enhancement of ET-1 vasoconstriction of aorta treated with SOD alone above that of aorta treated with ET-1 only. FIGURE

FIGURE 5. Effects of age on AP enhancement. x axis = dose of ET-1; y axis = mean difference in % contraction/lOO; z axis = treatment.

CRAWFORD el al.: VASOACTMTY OF PEPTIDES 43

FIGURE 6. Effects of SOD on AD enhancement. x axis = dose of ET-1; y axis = mean difference in % contraction/100; z axis = treatment.

7 shows the minimal effect of the calcium channel blocker, verapamil, on the enhancement of vasoconstriction by ET-1.

To test whether endothelium was essential for the enhancement of vasoconstriction by AD peptides, endothelium was removed as described, and FIGURE 8 shows that enhancement does occur in the absence of endothelium. A comparison of controls to AP-treated aortas with and without endothelium suggests that the

FIGURE 7. Effects of verapamil on AD enhancement. x axis = dose of ET-1; y axis = mean difference in % contraction/100, z axis = treatment.


0

0

L

From left to right: .=no endothelium controls ( ~ 8 ) ;

(n=8); m=endothelium controls (~4); .=endothelium AP (n=8). ET-1 dose: ~ X I O - ~ M

FIGURE 8. Effects of endothelial removal on AD enhancement. x axis = dose of ET-1; y axis = mean difference in % contractioni100; z axis = treatment.

enhancement is greatest in the presence of the endothelium. In the absence of endothelium, FIGURE 9 shows that SOD has no effect on the enhancement response, but FIGIJRE 10 shows that verapamil can completely abolish the AP enhancement, without blocking the ET-1 contraction itself.

DISCUSSION

Our original findings of intrinsic vasoactivity and opposition to ACH-induced relaxation have been confirmed by Longmore et aL3 We have extended these

0

From left to right: .=Apalone ( ~ 8 ) ;

n=SODalone (n=6); .=ControB @=lo). ET-I dose: ~ x I O - ~ M

FIGURE 9. Effects of SOD on Ap enhancement in the absence of endothelium. x axis = dose of ET-1; y axis = mean difference in % contraction/100, z axis = treatment.

CRAWFORD et al.: VASOACTMTY OF PEPTIDES 45

FIGURE 10. Effects of verapamil on Ap enhancement without endothelium. x axis = dose of ET-1; y axis = mean difference in % contraction/100; z axis = treatment.

findings to enhancement of constriction of a second vasoconstrictor, endothelin. With this system we have explored the possible mechanism of vasoactivity. We demonstrate that the dose, timing, and activity of peptide fragments strongly suggests that the mechanisms of cytotoxicity is different from vasoactivity of AP peptides. The fact that Aj3 enhancement of vasoactivity occurs within seconds and minutes, and that A&, is vasoactive but is not, suggests that aggregation of Aj3 is not a prerequisite for vasoactivity. This contrasts with endothelial cytotoxicity where more amyloidogenic proteins are more toxic and require time to aggregate. Similar findings are noted in the neurotoxicity literature." The impact of age on the enhancement of vasoconstriction may be relevant when we consider the pathologic action of Aj3 peptides, which are particularly associated with the aging process.

In addition to our original conclusions regarding the role of the endothelium in Aj3-induced vasoconstriction and the relevance of free radical-mediated vasoactivity, the enhancement of vasoconstriction in the endothelin system observed with and without endothelium allows us to deduce that:

1. Aj3 causes the enhancement of vasoconstriction by a pathway that finally results in increased intracellular calcium levels in smooth muscle cells. The data do not preclude the possibility that Aj3 directly increases intracellular calcium levels in smooth muscle cells by opening calcium channels, stimulating receptors or by otherwise altering the smooth muscle cell membrane.

2. Raised intracellular calcium levels in endothelial cells are likely to result in nitric oxide synthase activation and the subsequent release of the vasodilator, NO. Thus in the presence of an intact endothelium, two forces would operate in the presence of AP peptides-an opposition to constriction provided by endothelial activity, and an opposition to relaxation provided by smooth muscle activation. In the absence of any forces that upset this balance, the addition of AP alone in an unperturbed system is likely not to result in any vasoactivity-a phenomenon we confirm in most isolated, untreated vessels that we observe. In the presence of Aj3, any compounds that result in changes in intracellular calcium levels are likely to emphasize any differential effect of Aj3 on the endothelial cells (relaxation) and smooth muscle cells (vasocon-


striction). Thus with endothelin, known to increase intracellular calcium levels, the overall effect is to increase constricting responses in the smooth muscle rather than the relaxing response in the endothelial cells. The blockage of calcium across both endothelial and smooth muscle membrane surfaces (i.e., by verapamil) results in little effect on blockage of vasoconstriction (FIG. 7) as relaxing responses are predicted to be blocked by verapamil also. However, in the absence of endothelial cells verapamil would be expected to block only the constricting responses, resulting in less vasoconstriction after AP. We thus conclude that calcium passage across cell membranes is a central action of AP in vasoconstriction, suggesting that this may be a key site for therapeutic intervention.

The question of whether cerebrovascular dysfunction and damage occurs in life by mechanisms similar to those described here in our isolated vessel experiments is unanswered. We have demonstrated that the apposition of soluble AP peptides with cellular components of the vasculature (particularly smooth muscle cells) can result in enhancement of vasoconstriction and the disruption of the normal balance of free radical-mediated vasotonicity. The pathophysiologic significance of these observations and their relationship to the neuropathology of AD are unknown. We regard the development of an animal model of cere- bromicrovascular dysfunction as the most direct method of determining whether AP peptides can cause cerebrovascular dysfunction, neuronal damage, and cog- nitive loss in life. To this end, a circulating serum AP peptide model is being developed in rats. Preliminary results show that chronically infused AP peptides can induce vascular damage (see the work of Su et al. in this volume) in rats, but the full animal model awaits further elaboration.

ACKNOWLEDGMENTS

This work was supported by the generosity of Mr. and Mrs. R. Roskamp. Our thanks are extended to Terrence Town for his assistance with formatting the figures.

REFERENCES

1. THOMAS, T., G. THOMAS, C. MCLENDON, T. SUTTON & M. MULLAN. 1996. P-Amyloid- mediated vasoactivity and vascular endothelial damage. Nature 380 168-171.

2. SUO, Z., C. FANG, F. CRAWFORD & M. MULLAN. 1997. Superoxide free radical and intracellular calcium mediate Afl,.42 induced endothelial toxicity. Brain Res. In press.

3. LONGMORE, J., T. GIOKARINI, L. BONAFINI, R. HILL & F. H E ~ I . 1997. P-amyloid-evoked changes in vascular reactivity are mediated via an endothelial-specific mechanism. This volume.

4. PIKE, C. J., D. BURDICK, A. J. WALENCEWICZ, C. G. GLABE & C. W. COTMAN. 1993. Neurodegeneratoin induced by beta-amyloid peptides in uitro: The role of peptide assembly state. J. Neurosci. 13: 1676-1687.

the vasoactivity of aβ peptides

Documents