 |
Description  |
|
|
BACKGROUND OF THE INVENTION
In the disease known as diabetes mellitus, the pancreas loses its ability
to manufacture and secrete insulin to counter rising blood glucose
concentrations, and metabolic imbalance results. Historically, the disease
has been treated by insulin injection, diet, exercise and, in limited
cases, oral medication. However, the treatment has been only marginally
successful at best.
A proposed solution is the artificial pancreas, a bedside device having
intravenous catheters for blood glucose sampling and insulin infusion, a
glucose analyzer and an insulin pump controlled by algorithms that take
into account information from the glucose analyzer. But there is a clear
need to develop a convenient miniature version that could be implanted to
provide the patient with totally automatic metabolic regulation. A
suitable insulin pump component for implantation on a clinical scale has
already been developed. But the glucose sensing component remains a
problem.
An electrolytic glucose sensor capable of assaying glucose in complex
fluids such as blood that has been removed from the body and exposed to
the atmosphere is disclosed in U.S. Pat. No. 3,542,662 to Hicks et al,
dated Nov. 24, 1970. In this sensor, an enzyme-containing membrane is
disposed between a fluid being assayed and a first oxygen sensor electrode
and a similar membrane not containing enzyme is disposed between the fluid
and a second reference oxygen sensor electrode.
In this sensor, a certain portion of the oxygen diffusing through the
enzyme-containing membrane is consumed by equimolar reaction with glucose
catalyzed by the enzyme and is therefore unavailable for detection by the
first oxygen sensor electrode. The second, reference oxygen sensor
electrode in which the membrane does not include enzyme, determines the
concentration of oxygen that would have been detected had not the
enzyme-promoted reaction occurred. And the difference in oxygen detected
by the two electrodes is relied on as proportional to the glucose
concentration.
However, when efforts were made to adapt this device for use directly in
the body, confusing observations resulted. The sensors were calibrated for
glucose response in the atmosphere before implantation in test animals.
After a period of implantation, glucose or insulin was injected
intravenously to perturb the blood glucose concentration. In the cases
where there was any response to glucose, the sensor-indicated
concentration was much less than expected.
The departure from expected behavior has been attributed universally to
unfavorable biocompatibility of the implant material. That is, implants
cause the development of an encapsulating sheath which, if thick and
dense, may be impermeable to glucose. However, this can be minimized by
the use of appropriate materials and does not adequately account for the
results secured.
A key problem is that the glucose concentration in the body is normally
higher than the oxygen concentration by a factor of 50 to 100 times. Since
the enzyme reaction is limited by the least abundant reactant, an
implanted sensor would respond to oxygen concentration rather than to
glucose concentration and therefor would be ineffective to measure glucose
concentration.
One proposal to solve this problem presented in an article by Fischer and
Abel entitled "A Membrane Combination for Implantable Glucose Sensors,
Measurements in Undiluted Biological Fluids" in Trans. Am. Soc. Artif.
Intern. Organs, Vol. XXVIII, 1982, involved sandwich membranes for
association with an oxygen electrode sensor. In these membranes, a
hydrophobic layer was disposed to cover an enzyme layer and a minute hole
directly aligned with the anode of the sensor was provided in the
hydrophobic layer to allow access of glucose from blood being assayed
without dilution to the enzyme and anode. Oxygen diffusing through the
surface of the entire hydrophobic layer was indicated to provide a
stoichiometric excess over the glucose.
The space and time within which reaction between glucose and oxygen must
occur in this arrangement are so limited as to impose limits on the range
of concentrations of glucose with which complete reaction can occur and
could affect the reliability of results obtained. Additionally, the small
amount of enzyme disposed for action on glucose entering the minute hole
tends to become inactivated in a relatively short time.
BRIEF STATEMENT OF THE INVENTION
The assay method of the present invention operates to determine the
concentration of a first component present in solution in stoichiometric
excess over a second component reactive with the first component by
diffusing the solution through a special membrane which reduces the rate
at which the first component in the solution passes through the membrane
to a value which is a function of the initial concentration, increases the
concentration of the second component so that the second component is
present in stoichiometric excess over the first component in the solution
which has passed through the membrane. After reaction of the first and
second component, the amount of unreacted second component or the amount
of reaction product is measured to determine the concentration of the
proportion of first component which has passed through the membrane.
The membrane of the present invention is a combination of a body of
hydrophilic material, which contains immobilized enzymes or catalyst for
promoting reaction between components from a body of an aqueous solution
and which provides a path through the membrane to a detector for dissolved
components from an aqueous solution of the components accessible to a
surface of the membrane, and a body or bodies of hydrophobic material
permeable to some but not all of the dissolved components disposed
relative to the hydrophilic material to effect a controlled change in the
relative proportions of the dissolved components from the proportions
existing in the original solution. In the membrane, the enzymes are
disposed in quantity and relation for long life and effective action on
the components, and the hydrophilic and hydrophobic bodies are arranged to
provide space and time for complete reaction of the components.
DRAWINGS
Reference is made to the drawings illustrating various embodiments of the
invention in which:
FIG. 1 is a schematic diagram of a sensor for operation according to the
present and employing the membrane of the present invention;
FIG. 2 is a cross-section on an enlarged scale illustrating one form of the
membrane of the present invention;
FIG. 3 is a cross-section on an enlarged scale illustrating a second form
of the membrane of the present invention;
FIG. 4 is a cross-section of an enlarged scale of a further form of the
membrane of the present invention; and
FIG. 5 is a cross-section on an enlarged scale of a still further form of
the membrane of the present invention.
DETAILED DISCLOSURE OF THE INVENTION
A sensor 10 incorporating the membrane 12 of the present invention is shown
(FIG. 1) with a surface 14 of the membrane 12 accessible to an aqueous
body 16 containing dissolved components of which the molecules of one
component are much larger or less soluble than the molecules of a second
component, and with the other membrane surface 18 adjacent a detector 20
for the second component. The membrane 12 is an association of a body of
hydrophilic material and a body or bodies of hydrophobic material arranged
in relations, such as shown in FIGS. 2 to 5, which may restrict the rate
at which the larger molecule component enters and passes through the
membrane 12 and/or may increase the rate at which the second component
enters and passes through the membrane.
The hydrophilic material of the membrane is permeable to both a large
molecule component such as glucose and a small molecule components, such
as oxygen, in the solution and is disposed to provide a path through the
membrane from the body of solution being assayed.
An enzyme or a catalyst for promoting the reaction between the large and
small molecule components may be immobilized in the hydrophilic material
for action on these components as they diffuse through it. Or
alternatively a separate enzyme-, or catalyst-containing component may be
provided for action on the components that have passed the membrane.
The hydrophobic material of the membrane is impermeable to the larger or
less soluble molecule component but permeable to the smaller or more
soluble molecule. This hydrophobic material may be disposed in a body or
bodies which limit the surface area of hydrophilic material exposed for
accepting large molecule components from the solution and thus reduce the
rate of entry of such components to a value which is a function of the
concentration existing in the solution, so that the rate of component
entering is that which would enter from a more dilute solution in the
absence of the hydrophobic material. Additionally, substantial surface
area of hydrophobic material is provided for accepting the small molecule
component. Alternatively, the hydrophobic component may be dispersed as
small domains in a continuous phase of the hydrophilic material to reduce
the front along which the large molecule component can be transported and
so reduce its effective diffusion coefficient or transport rate, while the
small molecule material can diffuse at a high rate since it can move
through both the hydrophilic material and hydrophobic domains. The result
of limitation of the rate of entry and/or transport of the larger molecule
component and the increased rate of entry and transport of the smaller
molecule increases the ratio of small molecule material to large molecule
material passing through the membrane to the detector.
In the following disclosure, the invention will be referred to primarily in
the assay of glucose in solutions such as blood, containing a large
stoichiometric excess of glucose over oxygen; but it is to be understood
that the membrane and method of the invention are not limited to this use
or assay of this material, but may be used for the assay of other large
molecule compounds such as lactates, cholesterol and others in solution
containing oxygen, ammonia or other low molecular components reactive with
the larger molecule components. Also the sensor primarily referred to is
electrolytic and includes an oxygen-sensitive electrode detector which may
be similar to the oxygen-sensitive electrode detectors of the Hicks et al
patent referred to above; but non-electrolytic sensors using optical, e.g.
glass fiber optics, or other known detectors may be used.
A parameter used as a first approximation to show the relationship between
factors affecting entry and transport of components in the membrane is
##EQU1##
where .alpha. is the partition coefficient, D is the diffusion
coefficient, c is the concentration, .gamma..sub.o is a stoichiometry
coefficient that may range from 0.5 to 1.0, the subscripts o and g refer
to oxygen and glucose respectively, and the subscript B refers to the
concentration of that solute in the well-mixed bulk phase of the sample.
Other factors that affect the ratio of substrates in the membrane are: the
external mass transfer resistance, the enzymatic reaction, and whether or
not the sensing electrode consumes the oxygen that it senses. These latter
factors are important to the operation of the sensor, but are irrelevant
to this discussion. When the value c*.sub.o of is greater than or equal to
unity, there will be adequate oxygen in the membrane and the system will
be glucose-limited and capable of assaying glucose. Thus, from the
anticipated bulk glucose and oxygen concentrations that the sensor will be
exposed to, the ratio of the parameters to maintain c*.sub.o .gtoreq.1 and
reliably determine glucose may be determined.
The diffusion coefficient D indicates how fast the respective solute can
move once within the membrane. A smaller solute such as oxygen can diffuse
faster in most materials than a larger molecule like glucose. In
membranes, the rate of diffusion is determined by the internal microscopic
structure, the degree of crosslinking of the network or "tightness", and
the flexibility of the polymeric chains composing the membrane. In
principle, increasing the crosslink density of a membrane should be
favorable for alleviating the oxygen deficient, since this would slow the
rate of glucose diffusion to a greater extent than it would affect oxygen.
However, in practice, it is difficut to crosslink hydrophilic gel
membranes to that extent that the differential retardation of glucose
could achieve the differences needed here.
The partition coefficient .alpha. indicates the solubility of the solute in
the membrane and is determined by the type of bonding that can exist
between the solute and the membrane material. For example, a hydrophilic
or "water loving" solute such as glucose is readily partitioned into a
hydrophilic, gel-type material, but is excluded from a hydrophobic
material. The enzymes are hydrophilic and also partition into this type of
material. Oxygen dissolves to some extent in water and therefore can also
partition into a hydrophilic material, but in addition is soluble in
certain hydrophobic materials. Oxygen is, in fact, highly soluble and
diffusible in certain hydrophobic materials such as silastic rubber. This
feature is used here. Changing the membrane thickness is not advantageous
since it affects both solutes to the same extent.
In the membrane embodiment shown in FIG. 2, the membrane 12 is formed with
a continuous phase 22 of hydrophilic material in which very small bodies
or domains 24 of hydrophobic material are distributed. Hydrophilic
materials useful in the continuous phase include polyacrylamide gel,
glutaraldehyde-crosslinked collagen, polyhydroxyethylmethacrylate and its
derivatives and other hydrophilic polymers and copolymers. A hydrophilic
polymer can also be made by crosslinking oxidase or other enzymes with
glutaraldehyde or other crosslinking agents.
Hydrophobic materials for association with the continuous phase 22 of
hydrophilic material include polydimethylsiloxane, polymers of
tetrafluoroethylene or its fluoro-chloro analogues alone or as copolymers
with ethylene or propylene, polyethylene, polypropylene, cellulose acetate
and other oxygen-imbibing polymeric materials. These materials will be
included in amount of from about 2% to about 40% of the weight of the
membrane.
A number of procedures are available for combination of the hydrophobic
material with the hydrophilic material. The hydrophobic material in solid
form may be fractured or milled into miron particle size, in the range of
from about 1 to about 10 microns, treated with a surfactant such as
gamma-aminopropyltriethoxysilane and mixed with fluid uncrosslinked or
unreacted, e.g. monomeric or partially polymerized hydrophilic phase
material preferably containing enzyme or catalyst material. The mixture is
subsequently cast as a thin layer and crosslinked or polymerized to
solidify it. Or the hydrophobic material may be in the form of a latex of
which the small particles of hydrophobic material already carry surfactant
and are suspended in water so that the material is readily mixed with the
fluid hydrophilic material.
In another procedure, a viscous polymeric hydrophobic material is mixed
with or entrapped in the hydrophilic material. For example, viscous
polydimethylsiloxane of various molecular weights have been mixed with
collagen or with polyacrylamide and cast as a membrane.
In an alternative arrangement (see FIG. 3), the hydrophobic material is a
highly porous film or solid 26 which is treated with a surfactant, filled
with fluid hydrophilic material 28 and treated to crosslink or polymerize
the hydrophilic material in place to form an interpenetrating network
structure.
Another form of membrane (see FIG. 4) disposes the hydrophobic material as
a stratum 30 covering a hydrophilic stratum 32 and formed with spaced
small openings 34, having a size and number to control the entry of the
large molecule substrate into the hydrophilic stratum 32 while providing a
broad surface for entry of the gas. In this form, glucose entering the
openings 34 diffuses through the stratum 32 between the hydrophobic
stratum 30 and the surface of the detector 33 generally parallel to the
face of the membrane toward the electrode 35 and reacts with oxygen
entering through the hydrophobic stratum 30 through the action of enzyme
or catalyst present in the stratum 32.
A further form of the membrane (see FIG. 5) disposes the hydrophilic
component as a thin, preferably immobilized enzyme-containing a layer 36
with a thin layer 38 of hydrophobic component covering a portion or all of
the face 40 of the hydrophilic layer 36 leaving the edge surfaces 42 due
to the thickness of that layer accessible to the solution.
In the membrane shown in FIG. 5, the large molecule material, such as
glucose, enters the membrane only through the edge surfaces 42 of the
hydrophilic layer 36 and diffuses toward the center portion substantially
parallel to the face of the membrane between the hydrophobic layer 38 and
the surface of the detector element 44 which may be an oxygen electrode.
The reactive gas, e.g. oxygen, canenter the membrane through the entire
exposed surface 46 of the hydrophobic layer 38 and insures that there is
an excess of oxygen over glucose.
This last form of membrane offers special advantages in enabling longer
enzyme activity. In the previous forms, there is need to make the membrane
as thin as possible and as the enzyme becomes inactive, the substrates can
effectively diffuse farther and farther into the membrane before
encountering active enzyme. Ultimately the substrates can diffuse all the
way across the membrane without reaction and a sensor would no longer
detect glucose.
But with the design of FIG. 5, advance of inactivation is effectively in
the direction of glucose diffusion parallel to the plane of the oxygen
electrode. The dimension in this direction can be made very long, thus
allowing a great reserve.
The configuration of this form can be made in a variety of geometrics in
addition to the one shown in FIG. 5. Thus, the oxygen electrode could be
made as a cylinder with a circumferential hydrophilic coating containing
enzymes and with a hydrophobic coating over the side of the hydrophilic
coating. In this configuration, glucose could enter only through the
hydrophilic exposed edge and oxygen could enter through the sides.
Still another arrangement is a combination of a circular plate oxygen
electrode with an enzyme-containing layer on each side and with a
hydrophobic layer on the outer side of this layer. This allows oxygen to
enter from both directions perpendicular to the plate and glucose to enter
from the circumferential edges.
In operation of an electrode sensor 10 (see FIG. 1) for determining the
concentration of glucose in a body 16 of a solution containing dissolved
glucose and oxygen using a membrane 12 containing immobilized glucose
oxidase and catalase, the oxidase enzyme promotes the following reaction:
glucose+O.sub.2 +H.sub.2 O.fwdarw.glucono-.delta.-lactone+H.sub.2 O.sub.2
and the lactone hydrolyzes spontaneously to gluconic acid. The H.sub.2
O.sub.2 may be decomposed by the catalase in the reaction:
H.sub.2 O.sub.2 .fwdarw.1/2O.sub.2 +H.sub.2 O
A certain portion of the oxygen diffusing through the membrane 12 is
consumed by equimolar reaction with glucose catalyzed by the enzyme and
therefor unavailable for detection by the oxygen sensor electrode 20. An
identical reference oxygen sensor electrode 38, employing a similar
membrane 40 which does not contain enzyme, determines the concentration of
oxygen that would have been detected had no enzyme reaction occurred. With
the appropriate sensor design, the difference in oxygen detected will be
proportional to the glucose concentration.
However, for the system to provide useful results, the glucose
concentration, as opposed to oxygen concentration, must be the limiting
factor. Since the reaction is limited stoichiometrically by whichever
component is present at the oxygen sensitive electrode in lowest
concentration, in order to make the system sensitive to glucose
concentration, oxygen must be present at the electrode in excess of the
glucose.
In the relation encountered in an implanted sensor, the normal
concentration of glucose in the blood, and presumably in the interstitial
fluid, is approximately 100 mg % or 5.5.times.10.sup.-3 M but may be much
higher on diabetics, while that of oxygen is likely to be 40 mm. Hg, or
only about 6.times.10.sup.-5 M.
In operation of the present sensor, according to the method of the present
invention, to insure response to glucose at tissue pO.sub.2, the enzyme
membrane 12 is designed so that oxygen passes readily into and through the
membrane and that a reduced amount of glucose, which is a function of the
concentration of glucose in the solution, diffuses into and through the
membrane 12 for reaction with the oxygen. The ratio of oxygen to glucose
is controlled by the entry and/or transport properties of the special
membrane so that the relative proportions are changed from a concentration
ratio in the body of liquid, e.g. blood, of between approximately 50 and
100 parts of glucose to one of oxygen to a new ratio in which a slight
stoichiometric excess of oxygen exists in the membrane. By this action,
the system is not stoichiometrically limited by the concentration of
oxygen in the body fluid and the system can operate on the simplifying
assumption that the reaction behaves as a one-substrate (glucose)
dependent process.
The following examples are given to aid in understanding the invention, but
it is to be understood that the invention is not limited to the particular
materials or procedures of the examples.
EXAMPLE 1.
Five grams of bovine collagen is suspended in 40 ml. of pH 7 aqueous
phosphate buffer and the suspension is heated in an autoclave at
100.degree. C. After removal from the autoclave and cooling, a clear
viscous PG,17 liquid was obtained.
A polytetrafluoroethylene latex (Teflon latex from DuPont) having a
particular size of 1 to 5 microns and a solids content of 10% was added in
proportions of 6.6 ml. of the latex to 3.3 ml. of the autoclave collagen
product, and 0.4 g. of glucose oxidase enzyme was added to the above
mixture and uniformly distributed therein.
The resultant mixture was cast on a glass surface as a layer of 10 mils wet
thickness and allowed to dry at room temperature for five hours. The dried
product was a film 3 mil in thickness. A 50% aqueous glutaraldehyde was
poured directly on the film and distributed uniformly over the surface.
The film was then rinsed to remove free glutaraldehye and peeled from the
glass surface.
A disk 1 cm. in diameter was cut from the film and edge bonded with
cyanoacrylate adhesive to the detector end of an electrolytic oxygen
sensor. The sensor was placed in a sealed vessel at known physiologic
oxygen concentration with the detector end immersed in physiologic buffer
solution; and the current due to the oxygen in the vessel was measured at
10 microamps.
Aliquots of glucose solution were added to the buffer solution in the
vessel in amounts to raise the glucose concentration by 1 mM for each
addition. The current due to glucose dependent oxygen was measured after
each addition and this was compared with the current generated by the
original buffer solution before glucose was added. The following table
lists the difference between these two currents after each aliquot
addition.
______________________________________
ALIQUOT ADDITIONS
0 1 2 3 4 5
______________________________________
Microamps 0 2 4 6 8 10
With Teflon
Microamps 0 4 8 10 10 10
Without Teflon
______________________________________
EXAMPLE 2.
A membrane is prepared as in Example 1, except that 0.4 g. of 1-lactate
oxidase obtained from M. phlei is immobilized in the mixture rather than
glucose oxidase.
A disc 1 cm. in diameter is cut from the film and edge bonded with
cyanoacrylate adhesive to the detector end of an electrolytic oxygen
sensor. The sensor is placed in a sealed vessel at known physiologic
oxygen concentration with the detector end immersed in physiologic buffer
solution. The current due to the oxygen in the vessel is measured at 10
microamps.
Aliquots of 1-lactic acid are added to the buffered solution in amount to
raise the 1-lactic acid concentration by 1 mM for each addition.
The current due to lactic acid dependent oxygen is measured after each
addition and is compared with the current generated in the original buffer
solution before addition of the lactic acid. The results are similar to
the results which were obtained in Example 1.
* * * * *
|
|
 |
|
 |
Description |
|
