Vesicle–enzyme communication
Fredric M. Menger,* Kingsley H. Nelson and Yingbo Guo
Department of Chemistry, Emory University, Atlanta, Georgia, 30322, USA
Acetylcholinesterase and chymotrypsin are able to catalyze
hydrolyses of vesicle-bound substrates at rates that depend
upon the ability of the substrates to project beyond the
membrane surface.
contact. It was the purpose of the work reported herein to learn
more about such enzyme–vesicle communication.
Substrates in our study consisted of acylated N-methyl-
6
7-hydroxyquinolinium iodide derivatives 2. Acetylcholinester-
Pathogenic cells have been identified that produce specific
enzymes in excessive amounts. For example, bone cancer has
been shown to exude unusually large quantities of alkaline
O
+
N
1
R
O
I–
phosphatase, while neuroblastomas generate high levels of
acetylcholinesterase.2 Earlier investigations from our labo-
ratory have exploited this phenomenon in connection with a
new and potentially specific mode of drug delivery involving
CH3
2
3
compound 1. There were two key reasons for synthesizing 1:
ase-catalyzed ester hydrolysis leads to a highly fluorescent
-hydroxyquinolinium salt (lex = 400 and 500 nm), allowing
kinetics to be carried out spectrofluorimetrically at substrate
(phosphate buffer; pH 6.0; 25.0
7
O
CH3
concentrations of only 5 m
M
C16H33O
N+ CH2CH2OAc Br–
°C), a concentration far too low to cause membrane disruption.
Substrates with hydrophobic R groups were incorporated into
the vesicle bilayers during their formation. Vesicles, consisting
of dilauroylphosphatidyl choline plus 5% dimyristoyl phospha-
C16H33O
CH3
O
1
7
tidic acid (to inhibit floculation), were prepared with the aid of
a LiposoFast low-pressure extruder equipped with a 0.1 mm
polycarbonate filter.8 Monodisperse unilamellar vesicles of
about 100 nm diameter (dynamic light scattering) at a total lipid
(
i) It possesses a pair of hydrophobic tails plus a charged
headgroup, the components essential for vesicle formation; (ii)
It is a substrate for acetylcholinesterase, an enzyme that
removes the acetate from the acetylcholine-like moiety. When
enzymatic hydrolysis occurs on the ‘vesiclized’ compound, the
freed hydroxy group reacts with one of the ester groups to form
a six-membered lactone, thereby eliminating a hydrocarbon tail.
But since amphiphiles with a single tail are generally unable to
maintain a bilayer structure, the vesicle ruptures and releases its
internal contents. The process is enzyme-specific, leading to the
possibility of selective release near a cancer cell that over-
produces the particular enzyme.
concentration of 1.7 m
M
were thereby achieved. The concentra-
tion of acetylcholinesterase (Sigma electric eel, type V-5) added
to the vesicular substrates was about 0.7 m as determined by
M
enzyme titration.9
In the absence of vesicles, but otherwise under the conditions
specified above, the acetyl form of 2 (called ‘C ’) reacts
2
instantly with acetylcholinesterase. Longer-tailed derivatives
are slower: the octanoyl, dodecanoyl, and tetradecanoyl esters
(‘C ’, ‘C12’ and ‘C14’) have half-lives varying from 20–60 s.
8
The hexadecanoyl ester (‘C16’) is not completely hydrolyzed
even after half an hour. Although micellization of C16 is not a
factor in the slow rate (its concentration lies well below the
critical micelle concentration), it is quite possible that chain-
coiling, as discussed in another context,10 sterically impedes
access to the headgroup by the enzyme.
The preceding mechanism presupposes a reaction between a
water-soluble enzyme and a vesicle-bound substrate. Somehow,
in a manner not yet understood, the substrate within the bilayer
membrane is able to find its way to the active site. It seems
unlikely that, prior to enzyme binding, a substrate migrates in
toto from the membrane into the aqueous domain. If this were
possible, vesicles would lose and exchange their water-
Introducing vesicles into the system had no effect upon the
fluorescence vs. time plots for C
concentrations (Fig. 1). Apparently, the hydrophilic substrate
fails to bind to the lipid bilayers. C also displays little rate
2
even at elevated phospholipid
4
insoluble lipids rapidly, which they do not. A reasonable
alternative is that substrates only partially extend themselves
beyond the membrane surface. Thus, a lipid molecule might
transiently expose its polar moiety (and even portions of its
hydrocarbon chains) to the bulk solvent. In this way an enzyme
would have an opportunity to grasp the substrate.
8
change when co-mixed with phospholipid, a result explainable
by either an absence of vesicle binding or by an efficient
enzyme catalysis on bound ester. NMR data, given below,
strongly favor the latter. In contrast, vesicular substrates
An X-ray structure of acetylcholinesterase (Torpedo cal-
ifornica) shows that the active site (containing a Ser-His-Glu
triad) lies at the bottom of a ‘deep and narrow gorge’. Hence,
C10–C16 experience substantial rate inhibitions, the magnitude
of which depend on the length of the chain (Fig. 1). Thus, C10
has a half-life of 300 s, whereas C14 and C16 are, for all practical
purposes, inert. The C14 and C16 tails likely serve to anchor the
substrates, i.e. to impede the ability of the headgroups to project
beyond the membrane surface where reactive encounters with
the enzyme’s cleft become possible. Note that the data in Fig. 1
have practical implications for prodrug design because release
of a membrane-bound moiety depends critically upon the
structure of the disposable addendum.11
5
a membrane-bound substrate must seemingly endure a tortuous
journey as it leaves the membrane and travels down the cleft to
the active site. In actual fact, the journey may not be as difficult
as it appears. The cleft is lined with aromatic amino acids that
are believed to guide into the active site a substrate molecule
reaching the outer rim of the cleft. If this is correct, then one
could visualize the escorting of a substrate molecule, which
happens to protrude from the membrane, to the active site as
soon as a vesicle and a properly oriented enzyme come into
NMR spectra in the absence and presence of vesicles leave no
doubt that C
8
(and, by inference, all substrate with chains longer
Chem. Commun., 1998 2001