Search Images Maps Play YouTube News Gmail Drive More »
Sign in

Patents

  1. Advanced Patent Search
Publication numberWO2006102762 A1
Publication typeApplication
Application numberPCT/CA2006/000487
Publication date5 Oct 2006
Filing date31 Mar 2006
Priority date1 Apr 2005
Publication numberPCT/2006/487, PCT/CA/2006/000487, PCT/CA/2006/00487, PCT/CA/6/000487, PCT/CA/6/00487, PCT/CA2006/000487, PCT/CA2006/00487, PCT/CA2006000487, PCT/CA200600487, PCT/CA6/000487, PCT/CA6/00487, PCT/CA6000487, PCT/CA600487, WO 2006/102762 A1, WO 2006102762 A1, WO 2006102762A1, WO-A1-2006102762, WO2006/102762A1, WO2006102762 A1, WO2006102762A1
InventorsTodd Hoare, Robert Pelton
ApplicantMcmaster University
Export CitationBiBTeX, EndNote, RefMan
External Links: Patentscope, Espacenet
Glucose responsive microgels
WO 2006102762 A1
Abstract
A functionalized microgel characterized by the presence of phenylboronic acid (PBA) functional groups is disclosed. A process for manufacturing the microgel, comprising the step of grafting aminophyenylboronic acid (APBA) to a carboxylated microgel via EDC coupling is also disclosed. The microgel can also be manufactured by polymerizing a monomer containing phenylboronic acid (PBA). The microgel can be adapted to deswell in the presence of glucose; to swell in the presence of glucose under human physiological conditions; to maintain a net negative surface charge at human physiological pH value; to, when exposed to glucose, either swell or contract in response to pH; or contain both anionic and cationic charges under physiological conditions. Uses of the microgel inter in water treatment, biological separation, glucose sensors, self-regulating insulin delivery devices and fluid switches are also disclosed.
Claims  (OCR text may contain errors)
1. A process for manufacturing functionalized microgels comprising the step of grafting aminophyenylboronic acid (APBA) to a carboxylated microgel.
2. A process according to claim 1, wherein the APBA is grafted to the carboxylated microgel via EDC coupling.
3. A process for manufacturing functionalized microgels comprising the step of copolymerizing a monomer containing phenylboronic acid (PBA).
4. The functionalized microgel when made by the process of any one of claims 1 to
3.
5. A process for manufacturing amphoteric functionalized microgels comprising the step of copolymerizing a cationic monomer and an anionic monomer using a cationic surfactant and an anionic initiator.
6. The amphoteric microgel when made by the process of claim 5.
7. A functionalized microgel characterized by the presence of phenylboronic acid (PBA) functional groups and further characterized in that it deswells in the presence of glucose.
8. A functionalized microgel characterized by the presence of phenylboronic acid (PBA) functional groups and further characterized in that it swells in the presence of glucose under human physiological conditions.
9. A functionalized microgel characterized by the presence of phenylboronic acid (PBA) functional groups and further characterized in that a net negative surface charge is maintained at human physiological pH values. 10. A functionalized microgel charcterized by the presence of phenylboronic acid (PBA) functional groups and further characterized in that, when exposed to glucose, it either swells or contracts in response to pH.
11. A functionalized microgel characterized by the presence of phenylboronic acid (PBA) functional groups and further characterized in that it contains both anionic and cationic charges under physiological conditions.
12. A process for the production of potable water from a water supply containing biological material, comprising the step of: adding a functionalized microgel characterized by the presence of phenylboronic acid (PBA) functional groups to said water supply to produce an aqueous mixture containing a biological material/microgel floe; and extracting the floe from the aqueous mixture to leave said potable water.
13. A process according to claim 12, comprising the further step of removing the functionalized microgel from the floe for reuse by increasing the temperature and/or decreasing the pH.
14. A process for measuring the concentration of glucose in solutions, comprising the steps of: adding a functionalized microgel characterized by the presence of phenylboronic acid (PBA) functional groups to said solution; and monitoring the particle size and/or the light scattered by the microgels.
15. A process for separating glucose-consuming microbes at a specific phase in then- growth cycle from a growth medium including a glucose nutrient, comprising the step of: providing a mixture of said growth medium, inoculated with glucose-consuming microbes, and of a functionalized microgel characterized by the presence of phenylboronic acid (PBA) functional groups and further characterized by a tendency to flocculate said microbes when glucose concentration falls beneath a threshold level
- 48 -
SUBSTiTUTE SHEET (PaE 26) which level, in the context of said growth medium, corresponds to said specific phase in the growth cycle of said microbes; and extracting the floe produced.
16. A separation process for use with a cell mixture containing target cells and collateral cells having different surface chemistry properties, comprising the steps of: adding to said mixture a functionalized microgel characterized by the presence of phenylboronic acid (PBA) functional groups and further characterized by a tendency to selectively flocculate said target cells and to refrain from floe formation with said collateral cells; and extracting the floe.
17. A sensor for detecting glucose in liquids, said sensor comprising: a probe containing a functionalized microgel characterized by the presence of phenylboronic acid (PBA) functional groups and being further characterized in that it undergoes a volumetric change in response to changes in glucose concentration, said probe being adapted to expose said microgel to said liquids; a light source for directing light to said microgel; and a meter for detecting the difference in intensity between the light entering said microgel and the light leaving said microgel.
18. A sensor for detecting glucose in liquids, said sensor comprising: a probe containing a functionalized microgel characterized by the presence of phenylboronic acid (PBA) functional groups and being further characterized in that it undergoes a volumetric change in response to changes in glucose concentration, said probe being adapted to expose said microgel to said liquids; a light source for directing light to said microgel; and a meter for detecting the colour of light diffracted from said microgel. 19. An insulin-delivery device comprising: a microgel network characterized by the presence of phenylboronic acid (PBA) functional groups; and a supply of insulin contained in said network, said network being characterized by a pore size which reversibly varies between: a smaller size, at relatively lower glucose concentrations, whereat a rate of insulin release from said network is relatively low; and a higher size, at relatively higher glucose concentrations, whereat the rate of insulin release from said network is relatively high.
20. An insulin-delivery device comprising: a microgel characterized by the presence of phenylboronic acid (PBA) functional groups; and a supply of insulin disposed in said microgel, said microgel being characterized by a charge which reversibly varies between: a net cationic charge when the level of glucose is relatively low, such that insulin is relatively tightly bound by electrostatic attraction to provide a relatively low rate of insulin release from said microgel; and a net anionic charge when the level of glucose is relatively high, such that insulin is relatively strongly repelled by electrostatic repulsion from said microgel to provide a relatively high rate of insulin release from said microgel.
21. Use ofthe functionalized microgel of any one of claims 4, 6, 7, 8, 9, lO and ll as a switch in a fluid circuit.
22. A vehicle to carry a drug to a target cell, comprising: a functionalized microgel network characterized by the presence of phenylboronic acid (PBA) functional groups, further characterized by an affinity to bind with said target cell and yet further characterized by an ability to contain said drug and release same at said cell. 23. A device for binding and detecting specific analytes, comprising: a functionalized microgel network characterized by the presence of phenylboronic acid (PBA) functional groups, and further characterized by an affinity to bind with said specific analytes.
- 51 -
SUBSTITUTE SHEET PXE 26)
Description  (OCR text may contain errors)

GLUCOSE RESPONSIVE MICROGELS

FIELD OF THE INVENTION

The present invention relates to microgels.

BACKGROUND OF THE INVENTION

Microgels are intermediates between linear polymers and hydrogels. Like hydrogels, microgels have a covalently-crosslinked, three-dimensional structure and can swell and deswell in response to changes in the gel environment; like linear polymers, they are relatively small in size (<1 micron) and consequently have very high specific surface areas, respond quickly to environmental stimuli, and can be made to flow.

SUMMARY OF THE INVENTION

The present invention comprises a technique of grafting aminophenylboronic acid to a pre-existing carboxylated microgel via EDC coupling to make PBA-functionalized microgels. Through the technique, very well-defined carboxylated microgels with different radial and chain functional group distributions can be synthesized. The PBA- functionalized microgels can exhibit controllable and predictable swelling responses. The microgels can also be produced by copolymerization of a PBA-containing monomer.

Super-responsive microgels can be designed with a volume phase transition temperature (VPTT) just below physiological temperature in the absence of glucose and just above physiological temperature in the presence of glucose.

m this use, the PNIPAM thermal transition is used to magnify the PBA-glucose interaction-induced transition within the microgels to achieve higher degrees of swelling response to a particular glucose concentration. Similarly, the temperature at which the transition occurs can be modified by adjusting the -COOH content of the microgel, the

- 1 -

SUBSTiTUTE SHEET (RULE M) degree of conjugation, and/or the distribution of functional groups in the microgel. This permits precise tuning of the temperature and/or glucose concentration at which large responses are desired as well as allowing the switch to be effectively turned "on" and "off' simply by changing the temperature.

The use of an amphoteric microgel allows (1) the design of microgels which are responsive to changes in glucose concentration at physiological pH (7.4) via the incorporation of cationic groups into the microgel; (2) a net negative charge to be maintained on the surface of the PBA-grafted microgel at all physiologically-relevant pH values, significantly reducing the risk of an immune response; (3) the exploitation of the anti-polyelectrolyte behaviour of amphoteric microgels to achieve significantly higher stability at high ionic strengths, permitting glucose-induced swelling under physiological conditions (pH 7.4, 0.15M ionic strength, and 370C); and (4) the synthesis of microgels which can contract in the presence of glucose as well as microgels which can switch their response between contraction and swelling according to the pH value of the medium (via regulation of the positive-negative charge balance within the microgel).

Using these microgels, microgel probes can be synthesized for measuring glucose concentrations in complex mixtures. The probes can be readily tuned (both in terms of the magnitude as well as the direction of the desired response) based on the temperature, pH, and salt concentration of the medium used. The anionically-initiated microgels are also expected to be ideal for use in self-regulating in vivo insulin control devices given their unique ability to swell up to 70% by volume under physiological pH, temperature, and ionic strength.

The present invention should also permit the construction of radiolabeled biotracking devices which can bind and detect specific analytes.

The microgels prepared can bind selectively to glucose and can be made to swell (or deswell, depending on the type of delivery envisioned) systematically with increases in the glucose concentration within the physiologically-relevant glucose concentration range

- 2 -

8ϋBSTlTUTE SHEET {B!JLE 2$ (0-3mg/mL). By loading insulin into the microgels, it should be possible to provide for controlled release rates of insulin, into the bloodstream based on the blood glucose level. As insulin is released, the blood glucose level will drop and the gel will deswell, reducing the diffusion rate of insulin out of the microgel and avoiding "overshoots" in blood glucose correction. Alternately, the observed inversion of the amphoteric microgel surface charge from positive to negative as the glucose concentration is increased at physiological pH may be applied to regulate glucose-sensitive insulin release. Given that the microgels are highly uniform and easily injectible (unlike macrogels which have previously been synthesized), such a system should be useful as a self-regulating diabetes treatment, reducing the frequency of insulin inj ections and eliminating the risk of insulin overdose and subsequent hypoglycemia.

Most living cells as well as many viruses are coated with a highly hydrated carbohydrate layer containing significant concentrations of the cis diol-containing carbohydrates which bind to phenylboronic acid. PBA-containing microgels with a low surface charge can selectively bind to these carbohydrates in the cell glycocalyx and viral envelope to selectively remove such materials from complex mixtures. This would allow, for example, water to be made "safe" by using a comparatively small amount of flocculant since very little of the polymer would be "wasted" in removing inert materials (ie. clays, sands, minerals) from the water. Also, given the temperature and pH sensitivity of the microgel, it would be possible to separate the microgel-cell/virus complex from the other components of the mixture (easily accomplished with microgels via filtration or centrifugation) and subsequently recover the biological component by increasing the temperature and/or decreasing the pH. In the former case, the large surface area reduction achieved in thermal deswelling may sterically induce desorption; in the latter case, since glucose only binds to ionized PBA groups, lowering the pH below the pKa of the boronic acid groups (ie. from pH 7.4 to pH 6, well within the "safe" pH range for handling cells) would eliminate the affinity of PBA groups for the carbohydrates and induce desorption. The microgel could also be recycled for multiple use if desired.

- 3 -

SUBSTITUTE SHEET (BULE 20) By monitoring either the particle size or the intensity of light scattered by the PBA- conjugated microgels, one can measure the concentration of glucose in solutions in situ without requiring complex separation/isolation procedures. Of particular significance, this detection can be performed in physiological ionic strength solutions using the present invention. Furthermore, by changing the temperature and/or pH of the medium, one can selectively turn the microgel detection system on or off as desired. Microgels with enhanced sensitivity over specific glucose concentration ranges can also be synthesized to provide higher resolution to detect smaller glucose concentration changes over ranges of interest. Furthermore, the glucose concentration range over which this additional sensitivity occurs can be tuned simply by changing the operating temperature of the microgel sensor.

The highly homogeneous, spherical nature of the PNIPAM-PBA microgels make them ideal for use in soft switches or as actuators in nano-fabricated arrays or fluidics setups. Given that microgels can be designed to both swell and deswell in the presence of glucose, persons of ordinary skill in the art using this disclosure could easily design a microgel switch which would close (swelling gel) or open (deswelling gel) as the glucose concentration is increased.

Further advantages, features and characteristics of the present invention will become more apparent upon consideration of the following detailed description and accompanying Figures, the latter being briefly described hereinafter.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1 shows ratio change in hydrodynamic volume (V, Fig. 8.1 (a)) and light scattering intensity (LSI, Fig. 8.1(b)) as a function of glucose concentration for the APBA-modified microgels at pH 9 and 250C (5mM ionic strength). The^-axis in both cases is the ratio between the variable value in an x mg/mL glucose solution (where x = 0.5mg/mL or lmg/rnL) and the variable value in a glucose-free solution. The

- 4 -

SUBSTITUTE SHEET (RULE 20) corresponding electrophoretic mobilities measured at Omg/mL (top) and lmg/mL (bottom) glucose concentrations are also shown above each microgel data in panel (a).

Figure 2 shows hydrodynamic versus temperature profiles for the APBA-AA-11 A5 E50 microgel in the presence of physiologically-relevant glucose concentrations at pH 9. The VPTT trend shown along the x-axis illustrates to the midpoint VPTT value corresponding to the transitions at each glucose concentration.

Figure 3 shows volumetric (Fig. 3(a)) and light scattering intensity (Fig. 3(b)) ratios for APBA-AA-11 A5 E50 as a function of glucose concentration at different temperatures (pH 9, 5mM carbonate buffer)

Figure 4. shows hydrodynamic diameter versus temperature profiles for PBA-microgel conjugates prepared using microgel platforms with different functional group distributions (pH 9, 5mM carbonate buffer). The filled points represent the size profile in the absence of glucose while the unfilled points show the diameters in the presence of lmg/mL glucose.

Figure 5 shows ratio change in hydrodynamic volume (lighter bars) and light scattering intensity (darker bars) and absolute change in electrophoretic mobility (text above/below bars) in a pH 7.4 5mM phosphate buffer for APBA-17-4 microgels prepared under different conjugation conditions when exposed to lmg/mL glucose

Figure 6 shows ratio change in hydrodynamic volume (light shaded bars) and light scattering intensity (dark shaded bars) and absolute change in electrophoretic mobility (text above bars) for cationic-initiated amphoteric microgels exposed to lmg/mL glucose at different pH values (5mM carbonate or phosphate buffers).

Figure 7 shows ratio change in hydrodynamic volume (blue bars) and light scattering intensity (purple bars) and absolute change in electrophoretic mobility (text above bars) for anionically-initiated amphoteric microgels exposed to lmg/mL glucose at different pH values (5mM carbonate or phosphate buffers).

Figure 8 shows size distributions of E. coli. suspensions upon the addition of the (a) AMPH(+)-17-5 A5 E50 or (b) AMPH(+)-17-5 A5 ElOO microgel upon the addition of various glucose concentrations (pH 7.4, 5OmM carbonate buffer, 25C). Size distributions are measured 10 minutes after microgel addition.

Figure 9 shows size distributions of E. coli. suspensions upon the addition of the cationic-initiated amphoteric microgel APBA(+)-17-5 in the absence of glucose (pH 7.4, 5OmM carbonate buffer, 25C). Size profiles are recorded one half hour after microgel addition.

Figure 10 shows size distributions of P. pastoris upon the addition of (a) APBA-22-0 A5 E20 and (b) APBA-22-0 A5 ElOO as a function of added glucose (pH 9, 5OmM carbonate buffer, 25C). Size profiles are recorded one half hour after microgel addition.

Figure 11 shows size distributions of P. pastoris suspension upon the addition of microgels containing different PBA contents in the absence of glucose (pH 9, 5OmM carbonate buffer, 250C).

Figure 12 shows cell-selective fiocculation using the anionic PBA-functionalized microgel APBA-22-0 A5 ElOO (pH 9, 5OmM carbonate buffer, 25C): (a) E. coli. response; (b) P. pastoris response

Figure 13 shows effect of the carbohydrate inhibiting agent on yeast fiocculation control: (a) particle size distributions; (b) laser obscuration values as a function of fiocculation time

Figure 14 shows effect of salt on glucose fiocculation control: (a) particle size distributions; (b) laser obscuration values as a function of fiocculation time Figure 15 shows effect of the pH on yeast flocculation control: (a) particle size distributions; (b) laser obscuration values as a function of flocculation time

Figure 16 shows hydrodynamic diameter and light scattering responses of a series of microgels with different total PBA contents (pH9, 250C, 5mM ionic strength): (a) light scattering intensity ratio (relative to zero-glucose); (b) light scattering intensity trend expanded for the APBA-22-0 and APBA-11-0 microgels in the 0-0.2mg/mL glucose range; (c) hydrodynamic diameter ratio (relative to zero-glucose); (d) hydrodynamic diameter trend expanded for the APBA-22-0 and APBA- 11-0 microgels in the 0- 0.2mg/mL glucose range

Figure 17 shows electrophoretic mobility versus pH profiles for APBA(-)-28-8 A2 E20 and APBA(-)-28-8 A5 E50 in the presence and absence of 2mg/mL glucose.

DETAILED DESCRIPTION

The production of various microgels and their physical properties are described hereinafter.

Materials: N-isopropylacrylamide (NIPAM, 99%, Acros Organics) was purified by recrystallization from a 60:40 toluene:hexane mixture. Methacrylic acid (MAA, 99%, Aldrich), acrylic acid (AA, 99%, Aldrich), and N,N-dimethylamine ethyl methacrylate (DMAEMA, 98%, Aldrich) were purified by vacuum distillation. N,N-methylene-bis- acrylamide (MBA, 99+%, Aldrich), vinylacetic acid (VAA, 97%, Aldrich), fumaric acid (FA, 99%, Aldrich), cetyltrimethylammonium bromide (CTAB, 95%, Aldrich), sodium dodecyl sulfate (SDS, 98%, Aldrich), 2,2'-azobis(2-methylpropionamindine) dihydrochloride (AMPA, 97%, Aldrich), ammonium persulfate (APS, 99%, BDH), N-3- dimethyl(aminopropyl)-N'-ethylcarbodiimide (EDC, commercial grade, Aldrich), and 3- aminophenylboronic acid (APBA, 98%, Aldrich) were all used as received. All water used in the synthesis and characterization was of Millipore Milli-Q grade. Microgel Preparation: Polymerizations were conducted in a 25OmL three-necked flask fitted with a condenser and a glass stirring rod with a Teflon paddle.

A series of anionic microgels with the same total carboxylic acid content was synthesized using four different functionalized co-monomers; the recipes used are given in Table 1. In this case, the amount of functional monomer (MAA, AA, FA, or VAA) used was determined such that each end-product microgel contains the same number of moles of - COOH groups per gram of gel (as measured by conductometric titration).

Table 1. Polymerization recipes and measured -COOH contents for equal-COOH microgels.

Microgel Moles in Feed -COOH in

NIPAM FM SDS MBA APS product

(mmol/g gel)

APBA-MAA-6.5 1.24 X lO"2 7.9 x 10"4 1.7 x lO"4 6.5 x 10-4 4.4 x 10"4 0.52 0.03

APBA-AA-6.5 1.24 x lO"2 7.9 x 10"4 1.7 x lO"4 6.5 x 10"4 4.4 X lO"4 0.53 0.03

APBA-FA-6.5 1.24 x 10"2 6.9 x 10"4 1.7 x lO"4 6.5 X lO-4 4.4 x 10"4 0.51 0.03

APBA-VAA-6.5 1.24 x lO"2 35 x IQ-4 1.7 x lO"4 6.5 x 10"4 4.4 x IQ-4 0.50 0.02

Four microgels with different acrylic acid contents and two microgels with different VAA contents were synthesized; the recipes used in those preparations and the final carboxylic acid content of the microgels (as measured by conductometric titration) are given in Table 2.

Table 2. Recipes for anionic microgels containing different carboxylic acid contents

Microgel Moles in Feed -COOH in

NIPAM FM SDS MBA APS product

(mmol/g gel)

APBA-AA-6.5 1.24 x 10"2 7.9 x 10"4 1.7 x 10"4 6.5 X lO"4 4.4 X lO-4 0.53 0.03

APBA-AA-I l 1.24 x lO-2 14 X lO"4 1.7 x lO"4 6.5 x 10"4 4.4 x 10-4 0.75 0.03

APBA-AA-14 1.24 x lO"2 H x lO"4 1.7 x 10"4 6.5 x 10"4 4.4 x 10-4 1.05 + 0.04

APBA-AA-22 1.24 x lO"2 28 x 10"4 2.7 x 10"4 1.3 X lO"4 4.4 X lO"4 1.34 + 0.03

APBA-VAA-6.5 1.24 x 10'2 35 x 10-4 1.7 x 10"4 6.5 x lO"4 4.4 X lO-4 0.50 + 0.02

APBA-VAA-8.0 1.24 x IQ-2 47 X lO"4 1.7 x lO"4 6.5 X lO"4 4.4 X lO-4 ' 0.65 + 0.03

A series of amphoteric microgels was synthesized using acrylic acid (AA) as the anionic monomer and N,N-dimethylamino ethylmethacrylate (DMAEMA) as the cationic monomer. The recipes used in the synthesis of these microgels are given in Table 3. Table 3. Cationically-stabilized amphoteric microgel preparation recipes

Microgel Moles in Feed

NIPAM AA DMAEA CTAB MBA AMPA

AMPH(+)-12-0 1.24 x lO-2 1.9 x lO-J 0 1.1 x lO"4 6.5 x 10"4 3.7 x 10"4

AMPH(+)-22-0 1.24 x lO-2 3.4 x lO-3 0 1.4 X lO-4 6.5 x 10"4 3.7 x 1O-4

AMPH(+)-22-2 1.24 x lO"2 3.4 x 10"3 2.5 x 10"4 1.1 x lO"4 6.5 x 10"4 3.7 x 10-4

AMPH(+)-17-5 1.24 x lO"2 2.6 x 10"3 6.4 x 10"4 0.3 x 10"4 6.5 x 10-4 3.7 x 10"4

AMPH(+)-22-5 1.24 x lO"2 3.4 x 10"3 6.4 x lO"4 0.3 X lO"4 9.O x IO-4 3.7 X lO"4

AMPH(+)-27-7 1.24 x IQ-2 4.3 x IQ-3 8.3 x 10"4 0.3 x IQ-4 9.0 x IQ-4 3.7 x 10"4 * Each gel is labeled with the code "AMPH(+)-;c-/' where x is the mole fraction of AA and y is the mole fraction of DMAEMA in the feed (both with respect to the NIPAM feed).

Anionically-stabilized and initiated amphoteric microgels were also synthesized. Table 4 gives the recipes used to prepare the anionically-stabilized amphoteric microgels.

Table 4. Anionically-stabilized amphoteric microgel preparation recipes

Microgel Moles in Feed

NIPAM AA DMAEA CTAB MBA APS

AMPH(-)-17-5-2 1.24 x lO"2 3.4 x W 6.4 x lO"4 0.7 x 10"4 2.6 X lO-4 4.4 X lO"4 AMPH(-)-17-5-l 1.24 x lO"2 3.4 x lO'3 6.4 X 10"4 0.7 x 10"4 1.3 X lO"4 4.4 x 10"4 AMPH(-)-23-6-2 1.24 x 10"2 4.2 x lO'3 1.1 x 10"3 OJ x lO"4 2.6 x 10-4 4.4 x lO"4 AMPH(-)-28-8-2 1.24 x lO'2 5.6 x lO'3 1.4 X lO"3 2.5 x 10"4 2.6 x 10"4 4.4 x IQ-4

* Each gel is labeled with the code "AMPH(-)-x-y-z" where x is the mole fraction of AA, y the mole fraction of DMAEMA, and z the mole fraction of MBA in the feed (all with respect to the NIPAM feed).

Microgels were also prepared using a mixed charge system of an anionic initiator and a cationic surfactant. In spite of charge-charge interaction problems during the synthetic step, this mixed anionic initiator/cationic surfactant system was observed to give much higher overall microgel yields (-90% compared to -50% with the cationic initiator system), improved colloidal stability, and better control over the particle size of the product microgels over a broader size range compared to the fully cationic system. This is a key finding for practical scale-up of the synthesis of these microgels. Anionic- stabilized microgels tend to be more stable in complex mixtures (since most biological colloids carry a net negative charge) and anionic initiators are less susceptible to chain transfer, permitting higher yields of somewhat larger particles to be achieved.

For each series of microgels synthesized, all monomers, crosslinkers, and surfactants are dissolved in 15OmL of water and heated to the polymerization temperature of 7O0C under

- 9 -

SUBSTITUTE SHEET (PAJIE 26) a nitrogen purge. After 30 minutes, the initiator was dissolved in 1OmL of water and injected into the reactor. Polymerizations were carried out for 20 hours under 200RPM mixing. After cooling, all microgels were purified by at least five cycles of ultracentrifugation (Beckman model L7-55 at variable speeds according to the density of the microgel being purified), decantation and redispersion in water until the supernatant conductivity was less than 5μS/cm. Given the high conversions observed in the mixed- charge synthesis strategy used in the preparation of the anionically-initiated microgels, dialysis could also be used for microgel purification. Microgels were lyophilized and stored dry at 40C.

APBA Conjugation: Phenylboronic acid functional groups were incorporated on to the microgels via a "graft to" approach. EDC was used as a zero-length crosslinker to conjugate 3-aminophenylboronic acid (APBA) to the carboxylic acid groups contained in the base microgels. All conjugations were performed by suspending 36mg of lyophilized microgel in 3.6mL of MES buffer (pH 4.7, 2OmM ionic strength) in a scintillation vial. APBA solutions were prepared at different concentrations (ranging from lmg/mL up to lOmg/mL) in the same MES buffer and 3.6mL of the desired concentration solution was added to the microgel suspension and mixed under gentle magnetic stirring at 15C. After ~30 minutes, 1.OmL of a freshly-prepared EDC solution (again varying in concentration between 5mg/mL and 50mg/mL according to the desired degree of conjugation) was immediately added to the vials. All conjugated microgels referred to herein are given the code APBA-x-y Ap Eq, where x is the AA feed mole fraction, y is the DMAEMA feed mole fraction, p is the concentration of the APBA solution added (in mg/mL) and q is the concentration of EDC solution added (in mg/mL). For example, to prepare the microgel AMPH-27-7 A5 E50, 3.6mL of a 5mg/mL APBA solution and

1.OmL of a 50mg/mL EDC solution would be added to 3.6mL of a 1 Omg/mL AMPH-27- 7 microgel suspension. Even the minimum APBA and EDC concentrations chosen were in at least 10-fold molar excess to the total number of -COOH groups present on each of the microgels used in this study. Each conjugation reaction was allowed to proceed for 2 hours, at which point the microgel conjugates are immediately ultracentrifuged four times to remove unreacted APBA and EDC. The microgels are stored in suspension at 40C. Particle Sizing: Particle sizing was performed by dynamic light scattering using a detector angle of 90. A Lexel 95 ion laser operating at a wavelength of 514nm and a power of 10OmW was used as the light source. Correlation data was analyzed using a BI- 9000AT digital autocorrelator, version 6.1 (Brookhaven Instruments Corporation) and the CONTIN statistical method was used to calculate the particle size distributions. All microgels discussed in this study were monodisperse. Samples were typically prepared in thoroughly cleaned vials by suspending a small quantity of lyophilized microgel in (1) 0.005M pH9 carbonate buffer or (2) 0.005M pH7.4 phosphate buffer such that the scattering intensity was between 100 and 250 kilocounts per second. Glucose was added to selected samples in physiologically-relevant concentrations. Physiological blood glucose levels of interest are within the range of 0-3 g/L, with 0.9-1.30 g/L (5.0-7.2 mM) being the "normal" blood glucose concentrations recommended for diabetics. Hypoglycemia occurs when blood glucose levels dip below 4mM (0.7g/L) while hyperglycemia occurs when blood glucose levels rise above 1 ImM (2.0g/L). A practical drug delivery vehicle or in vivo glucose sensor must show a significant response within this range.

The pH of each sample was measured to ensure the buffering capacity of the solutions was not exceeded in any case. At least five replicates were conducted for each sample; the experimental uncertainties represent the standard deviation of the replicate measurements. Borate buffer (0.005M, pH9) experiments were also performed to confirm that the microgel swelling was induced by glucose-specific interactions. Free borate competitively binds glucose, reducing the amount of glucose bound to the microgel and thus the swelling response observed. For APBA- VAA-8.0 in the presence of lmg/mL glucose, only a 28% increase in volume is observed in borate buffer while a 90% increase in volume is observed in carbonate buffer at the same ionic strength.

Electrophoretic Mobility: Electrophoretic mobility values were measured using a ZetaPlus analyzer (Brookhaven Instruments Corporation) operating in phase analysis light scattering mode. Samples were prepared as described earlier for dynamic light scattering. A total of 10 runs (each comprised of 15 cycles) were conducted; the experimental uncertainties represent the standard error of the mean of the replicate runs.

Titration: Simultaneous conductometric and potentiometric titrations were performed using a Burivar-I2 automatic buret (ManTech Associates) as described previously. Briefly, 0.050g of lyophilized microgel was suspended in 5OmL of filtered 10"3M KCl. Quantitative data was acquired using slow base-into-acid titrations (67min/unit pH) to assure complete equilibration between the aqueous and gel phases. Nuclear Magnetic Resonance: 1H NMR analysis was conducted using a Bruker 600MHz spectrometer. Lyophilized microgels were suspended in D2O (Cambridge Isotope Laboratories) to acquire the spectra. Phenylboronic acid contents of the PBA- microgel conjugates were measured using the integrated intensity ratio of the aromatic hydrogen peaks (δ 7.2-7.6ppm) and the -CH group peak from NIPAM (δ 3.9ppm).

Insulin Uptake'. Insulin solutions were made by dissolving 2mg/mL insulin in a 0.02M hydrochloric acid solution and adjusting the pH to 4 by back-titrating the solution with sodium hydroxide. Insulin uptake was tested in both pH 4 (citric acid) and pH 7.4 (phosphate) buffers with ionic strength 1OmM. The pH-corrected insulin solutions were diluted in the appropriate buffer to a concentration of lmg/mL and uptake solutions were prepared by mixing 4.5mL of the resulting solution with 0.5mL of a 2mg/mL microgel suspension. This procedure results in insulin concentrations of 0.9mg/mL and microgel concentrations of O.lmg/mL in the uptake testing conditions. The insulin-microgel solution was mixed on a rotary mixer overnight and then centrifuged to separate the microgel-insulin complexes from the soluble insulin. The bicinchoninic acid (BCA) assay was used to determine the concentration of residual insulin present in the centrifuge supernatant. 2ml of the BCA reagent (containing bicinchoninic acid, sodium carbonate, sodium tartrate, and sodium bicarbonate in 0. IM NaOH, final pH 11.25) and 0.04mL of a 4% copper (II) sulphate pentahydrate solution were added to 0.1ml of the supernatant. The solution was immediately vortexed for 5 seconds and placed on the rotary mixer for 2 hours at 250C. The absorbance of the resulting solution was measured at 562nm using a UV spectrophotometer (Beckman DU800). Reference solutions were also prepared by substituting O.lmL of buffer in place of the centrifuge supernatant, accounting for the light green coloration of the BCA assay solution itself. The absorbance readings from the supernatant solutions were subtracted from this blank reference to give a net absorbance reading. The readings were converted to concentrations using a calibration curve developed by applying the BCA assay strategy to insulin solutions of known concentration over the linear range of the assay (0.2-1. Omg protein/mL). Three replicate measurements are performed for each microgel to confirm the accuracy of the results.

Insulin Release: A series of microgels were loaded with insulin using the uptake procedure described above in a pH 4 citric acid buffer solution. After centrifugation, the supernatant is assayed for insulin uptake while the pelletized insulin-microgel complex is resuspended in pH 7.4 phosphate buffer (1OmM). To simulate physiological conditions, 0.15M NaCl solution was added to 2 test tubes containing the pelletized gel while 0.15M NaCl and 2mg/ml glucose was added to a second pair of tubes. The solutions were redispersed and placed hi a shaker bath overnight at 37C to simulate physiological release. The microgel phase was then removed via filtration through a 0.2μm syringe filter pre-incubated with bovine serum albumin to block insulin adhesion to the filter. This method gives uptake predictions within the experimental error of the uptake values measured using ultracentrifugation as the microgel separation technique. The filtrate solution is analyzed for insulin content using the BCA assay as outlined in section 9.2.2.

Cell Preparation: E. coli. cells are cultured by adding ImL of the JM-109 strain into 5OmL of LB broth (1% tryptone, 0.5% yeast extract, 1% sodium chloride, pH 7.0) in the presence of O.lmg/mL ampicillin. The cells were incubated under shaking (200RPM, 37C) until the optical density reached ~0.6mg/mL. At this point, 0.5mg of isopropyl thiogalactoside was added to the suspension and incubation was continued for an additional two hours. Cells were harvested and washed via four successive centrifugation cycles into 1OmM pH 7.4 phosphate buffer and stored in a lOmg/mL suspension at 40C until use. AU flocculation experiments were completed within four days of cell culturing to ensure cell viability during the experiment.

- 13 -

SUBSTITUTE SHEET fRlHE 26) Yeast cells were also cultured by inoculating ImL of the Pichia pastoris in 5OmL of YEPD medium (1% yeast extract, 2% peptone, 2% dextrose). Cells were incubated by shaking at 200RPM overnight at 30C. An aliquot of 50μL of the resulting seed cells is then transferred to 5OmL of YEPD medium and incubated in a shake flask for 24 hours at 30C. Cells were harvested, cleaned, and stored using the procedure previously described for E. coli. Type II Saccharomyces cerevisiae yeast cells were purchased from Sigma- Aldrich and suspended at lOmg/mL in a 1OmM pH 7.4 buffer prior to experiments.

Cell Flocculation Experiments: Cell flocculation experiments were performed using a Malvern Mastersizer 2000. In each experiment, 6OmL of buffer (pH 4 citrate, pH 9 carbonate or pH 10 carbonate, all with ionic strength 5OmM) was added to the instrument reservoir and pumped through the optical cell to align the instrument, hi a typical experiment, 0.5mL of the lOmg/mL cell suspension was added to the reservoir and pumped through the optical cell at a rate of ~4mL/s to measure the size distribution of the cell suspension prior to polymer or carbohydrate treatment. Beam obscurations for each cell type tested ranged between 5-15% at this concentration. Glucose, cellobiose, or sucrose was then added to the desired concentration via addition of a 300mg/mL sugar stock solution prepared in water. After a five minute wait time, the size distribution of the cells was again measured. A background salt concentration of 45OmM was also added in some experiments to test the effect of ionic strength on the selective flocculation process. Finally, 0.5mL of a lmg/mL microgel suspension was added to induce flocculation and the size distributions were measured at 10-30 minute intervals. This results in a mass ratio of 10: 1 dry cells:dry microgel for a typical flocculation experiment. The suspension was mixed gently under magnetic stirring to ensure floes circulatation throughout the measurement. Each size distribution measurement was repeated three times, with the reported results representing the average of the distributions generated.

In selected cases, samples were withdrawn at pre-defined intervals during the flocculation process for turbidity measurements using an HF Scientific DRT-15CE turbidmeter. Microgels scatter some light but in all cases account for <2% of the total laser obscuration and turbidity values. Furthermore, while the addition of glucose or sucrose

- 14 -

SUBSTITUTE SHEET (RHi F ?6) does provide a nutrient feedstock for cell growth and proliferation, the pH conditions used for the flocculation testing are typically not conducive to cell growth and the Mastersizer size profiles of both the yeast and E. coli. cells remain constant over at least a three-hour period in the presence of sugar. As a result, changes in turbidity and/or light obscuration can be related to the concentration of cells remaining in suspension.

Effect of Functional Group Distributions on PBA-microgel Properties

As Table 5 indicates, the chain and radial functional group distributions were found to strongly affect both the efficiency of the PBA conjugation as well as the degree to which the microgel properties change upon conjugation.

Table 5. Properties of PBA-conjugated microgels with the same total carboxylic acid content but different functional group distributions; the APBA:NIPAM molar ratio is measured in the conjugated microgels via 1H NMR while the given conjugation efficiency is the percentage of the total number of -COOH groups in the microgel (measured via conductometric titration) which react with APBA during the conjugation.

Microgel ABPA:NIPAM Conjugation ] Particle Size Electrophoretic Mobility

Molar Ratio Efficiency (%) (pH 7.4, 25C) (nm) (pH 7.4, 25C) (xlO 8 πrVVs)

(%) Before After % Before After %

Change Change

MAA-NIPAM 3 46.2 408 5 372 3 -9 2 -0.45 0.03 -0.45 0.01 0 + 1

AA-NIPAM 4.5 69.2 299 + 4 221 + 1 -26 1 -1.16 0.03 -0.64 0.03 -45 3

VAA-NIPAM 6.2 95.4 271 5 199 + 3 -27 2 -1.46 0.06 -0.81 0.02 -45 3

FA-NIPAM 4.8 ' 73.4 342 4 332 6 -3 3 -1.67 0.04 -1.12 0.01 -33 1

AA-NIPAM, 0.1 1.5 297 2 310 5 +3 2 -1.16 0.03 -1.07 0.03 -3 2 no EDC

Without intending to be bound by theory, the inventors conclude as follows.

VAA-NIPAM has both a highly surface-localized radial -COOH distribution and a well- isolated chain -COOH distribution; consequently, the conjugation reactions can proceed relatively unhindered by steric barriers, allowing for an extremely high conjugation efficiency to be achieved (close to 100% conjugation yield). This yield is significantly higher than those previously reported for microgels systems and illustrates the advantage of engineering the microgel morphology to maximize functional group accessibility in grafting applications.

While FA-NIPAM has an even more surface-localized radial -COOH distribution (as evidenced by its significantly higher electrophoretic mobility prior to APBA

- 15 -

SUBSTITUTE SHEET {RULE 26) conjugation), the diacid nature of the monomer presents steric inhibitions to the conjugation of a second APBA molecule to the same FA monomer residue. As a result, a significantly lower conjugation efficiency is achieved with FA-NIPAM. Interestingly, virtually no change in particle size is observed upon APBA conjugation to the FA- NIPAM microgel, implying that the electrostatic repulsion between the remaining, highly surface-localized -COOH groups is sufficient to prevent the large deswellings observed in the other microgels. This is a potential advantage of using diacid-containing monomers in microgel preparations such that one charged -COOH group is maintained at most diacid sites after any given conjugation reaction; as a result, the volume change observed on conjugation is minimized and the final particle size of the conjugate can more easily be controlled at the initial polymerization step of the synthesis.

The degree of conjugation achieved with AA-NIPAM is very similar to that achieved with FA-NIPAM. This suggests that the negative effect of the significantly lower degree of functional group surface localization in AA-NIPAM (as shown by the electrophoretic mobility measurements in Table 5) is offset by the much higher average separation of the -COOH groups within the polymer chains (AA is a monoacidic monomer). However, despite the similar PBA contents in APBA-AA-6.5 and APBA-FA-6.5, much larger deswelling is observed upon conjugation with AA-NIPAM since the remaining, unconjugated -COOH groups are much less localized than in FA-NIPAM and the local charge-charge repulsion contribution to the gel swelling becomes minimal.

MAA-NIPAM has a relatively blocky chain distribution and a core-localized radial distribution of -COOH groups. Not surprisingly then, MAA-NIPAM exhibited the lowest degree of conjugation, no detectable change in the electrophoretic mobility before and after the conjugation, and a minimal deswelling upon conjugation since the few carboxylic acid groups which do react with APBA are located in the more highly crosslinked core of the microgel in which swelling/deswelling is elastically more restricted. It should be noted that the covalent conjugation of APBA to the microgel as well as the efficiency of the microgel purification protocol used were both confirmed by using a control sample in which the APBA solution and microgel suspension were mixed but no EDC was added. As shown in Table 5, no APBA was detectable in the microgel by NMR after purification and no significant change was observed in either the particle size or mobility data before and after conjugation.

Based on the observed morphology differences, it is expected that the microgels should exhibit significantly different responses in the presence of glucose. Figure 1 presents the observed changes in particle size (Fig. Ia) and light scattering intensity (Fig. Ib) of the four different -COOH-containing microgel/APBA conjugates.

With the exception of the EDC-free control, all microgels display some responsiveness to even the smallest physiologically-relevant glucose concentration. The magnitude of the response is determined both by the amount of APBA grafted to the microgel and the underlying morphology of the microgel. This is most clearly shown by the responses of the MAA-NIPAM microgel. Despite the fact that APBA-MAA-6.5 has the least total amount of APBA conjugated (almost half of that conjugated to the APBA-FA-6.5 microgel), the reduction in light scattering intensity upon glucose addition in APBA- MAA-6.5 is more than double that observed in APBA-FA-6.5 at the same glucose concentration. This is likely a direct result of the presence of APBA in the more densely crosslinked core of the microgel in APBA-MAA-6.5. hi this case, glucose binding causes ionization and thus swelling in the densest region of the microgel which scatters the most light; thus, even the relatively small degree of swelling shown hi Figure l(b) results in a large change in the light scattering intensity. Conversely, the larger number of APBA residues grafted in the more lightly-crosslinked, near-surface region of APBA- FA-6.5 results in a significantly larger increase in the particle diameter in glucose- containing solutions but a relatively low decrease in the light scattering intensity given that the denser microgel core remains unswollen. APBA-AA-6.5, with its intermediate morphology between the FA and MAA-functionalized microgels, exhibits an intermediate response to these two more extreme cases in both the light scattering intensity and particle size data.

The responsiveness of the microgels can be regulated by changing both the glucose concentration as well as the -COOH content of the microgel. As evidence for the former claim, in all cases, the microgel swells more and scatters less light in l.Omg/mL glucose solutions compared to 0.5mg/mL glucose solutions (although the volume ratio error bars do in some cases overlap). Thus, regardless of the morphology, systematic and concentration-dependent responses can be achieved to changes in the external glucose concentration. To illustrate the latter claim, a VAA-NIPAM microgel with 8mol% VAA was synthesized and subjected to the same conjugation conditions. Not only was more APBA conjugated to the microgel (7.1% APBA:NIPAM ratio according to 1H NMR compared to 6.2% for APBA-VAA-6.5) but larger responses were observed in both the light scattering intensity (Fig. l(a)) and particle size (Fig. l(b)) measurements to changes in the solution glucose concentration. Thus, the choice of comononier allows one to select the relative responses of the light scattering intensity and particle size to fluctuations in the glucose concentration while the choice of -COOH content allows for selection of the absolute magnitude of these responses.

Morphology control to achieve temperature-sensitive glucose responses

The volume phase transition can be used to (a) enhance the responsiveness of the conjugate microgels to glucose and (b) tune the APBA-microgel "sensor" to respond differently at different target glucose concentrations.

Enhancing microgel responsiveness to glucose: As shown in Figure 2, PBA-containing microgels exhibit a VPTT shift in the presence of glucose (~ +4C for the APBA-AA-11 A5 E50 microgel shown).

At temperatures well above and well below the VPTT, minimal glucose sensitivity exists in this microgel; only a -16% increase in volume is observed when lmg/mL glucose is added at 15C and no sensitivity whatsoever is observed at T > 350C. Thus, in a non-

- 18 -

SUBSTiTUTE SHEET (RULE ?fl thermosensitive microgel, more PBA would have to be added into the microgel matrix to result in significant enough swelling to apply the gel in any practical application. However, in the volume phase transition range, the Donnan swelling achieved by shifting the PBA ionization equilibrium via PBA-glucose binding is strongly amplified by the thermal swelling resulting from the VPTT shift upon glucose binding. Consequently, in the volume phase transition range 250C < T < 300C, the microgel shown in Figure 2 undergoes very large volume changes (up to a 270% volume increase at 3O0C) in the presence of glucose which would not be possible simply relying on the standard PBA- induced swelling mechanism in hydrogels. A similar trend is observed in the light scattering intensity (LSI) profiles, a potentially useful observation for sensor applications.

Thermal sensitivity also allows one to tune the glucose concentration range over which significant swelling and/or light scattering intensity responses are observed. Figure 3 shows how the temperature can strongly affect not only the magnitude but also the trajectory of the scattering intensity and/or volume dependence on glucose concentration. Glucose concentrations are chosen to represent physiologically-relevant situations; 0.4g/L is the hypoglycemic limit, 0.7-1.3g/L is the normal glucose concentration in blood, and 2g/L is the hyperglycemic limit.

At 25C, the VPTT of the microgel in the absence of glucose lies below the test temperature while the VPTT at all glucose concentrations lies above the test temperature; consequently, significant swelling is observed when a small amount of glucose is added but higher glucose concentrations have a minimal impact on gel swelling since only the Donnan equilibrium mechanism induces swelling (ie. minimal thermal swelling occurs). In contrast, at 27.5C, the microgel VPTT at glucose concentrations <1.3g/L also lies below the test temperature; consequently, a large, thermally-assisted swelling is observed as the glucose concentration increases from 1.3g/L to 2g/L (interestingly, the range over which insulin release would be particularly desirable to avoid hyperglycemia). A similar trend is observed using 3O0C as the test temperature, particularly in the light scattering intensity profile (Fig. 3(b)). Conversely, by lowering the temperature just above (32.5C) or just below (22.5C) the transition range, the microgel is only minimally responsive to

- 19 -

SUBSTtTl)TE SHEET (RULE 26) changes in the glucose concentration, effectively "turning off' the sensing ability of the microgel (particularly at T VPTT). Thus, by changing the temperature at which a PBA-microgel conjugate is used, one can achieve different glucose response profiles providing more or less resolution to changes in glucose concentrations, both generally and within specific, tunable ranges. This is of interest in sensor applications, allowing for (a) the precise tuning of the sensor response by changing the test temperature and (b) regulation of the glucose response at a specific test temperature by controlling the morphology of the base microgel.

A variety of strategies can be applied to enhance glucose-driven swelling in PBA- functionalized microgels, both overall and within very specific temperature and/or glucose concentration ranges. In any functionalized microgel, the swelling response to a specific stimulus can be maximized by increasing the concentration of the environmentally-responsive functional groups (increasing the specific driving force for swelling) and/or decreasing the crosslink density (reducing the elastic resistance to swelling). Table 6 shows the volumetric swelling and light scattering responses for a 22mol% acrylic acid functionalized microgel containing lmol% crosslinker and different APBA/PBA contents. In comparison, the AA-functionalized microgel used as the grafting platform in Figure 1 contains 6.5mol% acrylic acid and 4.5mol% crosslinker. The higher the APBA or EDC concentrations used in the synthesis procedure, the more PBA groups are grafted to the microgel.

Table 6. Ratio changes in hydrodynamic volume and light scattering intensity in the presence of lmg/mL glucose at different temperatures in pH 9 5mM carbonate buffer .

Microgel Ratio (x(lmg/mL Glucose)/x(0mg/mL Glucose)

Λ; = hydrodynamic x = light scattering volume intensity

APBA-AA-22 A2 ElO 1.28 + 0.09 0.31 + 0.01

APBA-AA-22 A2 E20 1.54 + 0.07 0.19 + 0.01

APBA-AA-22 A5 E20 3.54 + 0.17 0.15 + 0.01

APBA-AA-22 A5 ElOO 28.5 0.20 0.06 + 0.01

Up to 30-fold increases in volume and 95% decreases in light scattering intensity are achieved in the presence of lmg/mL glucose when microgels with high -COOH contents (i.e. more PBA binding sites) and lower crosslink densities are used as the conjugation platform. The volume and light scattering changes scale with the PBA content of the microgel; the higher the APBA and EDC concentrations used, the greater the light scattering and hydrodynamic diameter changes upon glucose exposure.

The temperature and glucose concentration range over which the glucose-induced phase transition can be thermally enhanced can also be controlled exploiting the thermosensitive swelling transition of the microgels. Table 7 shows how the density of PBA grafts to a specific platform microgel can regulate the swelling response.

Table 7. Ratio changes in hydrodynamic volume and light scattering intensity (LSI) in the presence of lmg/mL glucose at different temperatures in pH 9 5mM carbonate buffer

Microgel Ratio [x(lmg/mL Glucose)/x(0mg/mL Glucose)]

25C T = 37C x = Volume X = LSI x = Volume Jt = LSI

APBA-AA-Il Al E5 1.26 + 0.15 0.81 0.01 1.37 0.09 0.82 0.01 APBA-AA-I l Al ElO 1.60 + 0.12 0.69 + 0.01 4.25 + 0.30 0.84 + 0.01 APBA-AA-I l A5 ElO 3.11 + 0.27 0.65 + 0.01 3.72 0.21 0.90 + 0.02 APBA-AA-I l A5 E20 7.65 0.33 0.56 + 0.01 1.26 + 0.13 1.05 + 0.03

The VPTT of the unmodified 1 lmol% AA-NIPAM microgel platform is ~60C, well above both test temperatures. As more APBA or EDC is added to the conjugation mixture, more APBA is grafted to the microgel, making the microgel more hydrophobic and reducing the VPTT. At 25C, the VPTT of each of the APBA-modified microgels lies above the test temperature, resulting in a systematic increase in gel swelling and decrease in the light scattering intensity in the presence of lmg/mL glucose. A similar trend exists at 37C for the two microgels with the lowest APBA content. However, the PBA contents of both the A5 ElO and A5 E20 microgels are sufficiently high such that the VPTT of these microgels lie close to (A5 ElO) or below (A5 E20) the 370C test temperature. As a result, the glucose response is restricted in the A5 ElO microgel and effectively "shut off' in the A5 E20 microgel by the thermal phase transition. Thus, the temperatures at which both thermal amplification and suppression of the glucose- triggered phase transition occurs can be tuned based on the PBA content of the conjugate.

Similar control over the VPTT can be exercised by adjusting the number of non- conjugated carboxylic acid functional groups in the microgel. The higher the number of residual COOH groups in the microgel, the higher the VPTT of the microgel. Thus, by increasing the carboxylic acid content of the platform microgel, conjugates can be prepared which contain the same number of PBA groups (and thus show the same ionization-based glucose response) but exhibit thermally-enhanced glucose swelling at higher temperatures. Indeed, while Table 7 indicates that glucose-induced swelling of the APBA-AA-11 A5 E20 microgel is suppressed by the thermal transition at 370C, the corresponding APBA- AA-22 A5 E20 microgel swells more than three-fold by volume under the same conditions owing to its higher excess -COOH concentration.

Figure 4 shows how the functional monomer used to prepare the base microgel can also be used to tune the nature of the glucose-induced swelling.

The effect of functional group distributions on glucose-induced microgel swelling can be illustrated by comparing the deswelling profiles of the APBA-AA-6.5 and APBA-FA-6.5 microgels, both of which have approximately the same PBA and residual -COOH content . Both microgels swell ~2-fold by volume when exposed to lmg/mL glucose at 150C. However, APBA-AA-6.5 shows a clear VPTT shift, facilitating a thermally- enhanced five-fold swelling upon exposure to lmg/mL glucose at 27.5C, and maintains at least some glucose swelling activity at temperatures as high as 350C. Conversely, APBA-FA-6.5 exhibits no significant VPTT shift, swells no more than two-fold by volume in the presence of lmg/mL glucose at any temperature, and loses all glucose activity at T > 300C.

Overall, persons of ordinary skill using tin's disclosure can tune the VPTT of the resulting microgel to either amplify or suppress the glucose-induced phase transition within specific glucose concentration and temperature ranges by controlling the content and distribution of graft sites in the platform microgel (via the polymerization recipe used) and the degree of APBA conjugation in the graft microgel (via the EDC/APBA concentration), hi addition, several different temperature switches can be prepared from the same platform microgel using different conjugation conditions in the graft-to approach described, facilitating the rapid synthesis of multi-temperature responsive microgel mixtures if desired.

Amphoteric Microgels for Physiological Glucose Responsiveness

The microgels described to this point show glucose sensitivity only within 1 unit of the pKa value of the phenylboronic acid functional entity (~8.9). To apply the microgel for in vivo glucose sensing and/or insulin delivery, activity is required at physiological pH (7.4), temperature (37C), and salt concentrations (tonicity equivalent to a 0.15MNaCl solution). It is believed that no previous glucose-sensing gels have been demonstrated to operate under such conditions. Herein, the applicability of amphoteric PBA-grafted microgels for achieving physiological activity will be discussed.

Grafting PBA to amphoteric microgels has the same general effect as described earlier for anionic microgels: as the EDC and, to a lesser extent, APBA concentrations are increased, the graft yield of PBA groups increases and more free -COOH groups are consumed. However, in the case of the amphoteric microgels, as more -COOH groups are consumed by PBA grafting, the surface charge on the microgel may switch from negative to positive at pH values close to the pKa of the PBA functional groups, hi this case, the ionization of PBA functional groups upon exposure of the microgels to glucose actually decreases the net charge density within the microgels (effectively "neutralizing" some of the cationic charges), causing the microgels to instead contract in the presence of glucose. This behaviour, which is believed to have never been previously reported, is illustrated in Figure 5 for the APBA(+)-17-4 microgel at physiological pH. This behaviour should be very useful in microfluidics applications or "smart" insulin-releasing membranes in which insulin can escape via the microgel pores in the collapsed, high- glucose state but is blocked (or at least severely diffusion limited) when the microgels are in the swollen, low-glucose state.

Again, without intending to be bound by theory, the inventors comment as follows with respect to this behaviour.

- 23 -

SUBSTiTTUTE SHEET (RULE 26) At low APBA/EDC concentrations (APBA(+)-17-4 Al ElO), few PBA residues are grafted to the microgel and the net surface charge of the gel in the absence of glucose (as estimated by the electrophoretic mobility data) remains negative; consequently, the gel swells, but only minimally, upon exposure to glucose. As the APBA/EDC concentrations are increased, more -COOH groups are consumed and, since only a fraction of the PBA residues are ionized at pH 7.4, the surface charge of the gel in the absence of glucose becomes positive. In the case of APB A(+)- 17-4 A2 ElO, the charge is only slightly positive such that exposure to lmg/mL glucose can ionize enough additional PBA groups to again switch the surface charge back to a net negative value, resulting in gel swelling. However, in the case of APBA(+)-17-4 A2 E20 and APBA(+)-17-4 A.5 E50, the net surface charge simply becomes less positive in the presence of glucose, resulting in significant microgel deswelling (roughly half by volume in the case of the APBA(+)-17-4 A5 E50 microgel). The light scattering intensity responds in exactly the opposite fashion to the microgel volume.

This switch in the behaviour of the microgel from swelling to deswelling in the presence of glucose can also be achieved by changing the pH and, consequently, the ratio of PBA groups which are ionized at any given glucose concentration. Figure 6 illustrates the swelling behaviour of three amphoteric PBA-containing microgels as measured at both pH 7.4 and pH 9.

In this case, two variables must be considered to understand the experimental data: the -COOH/-N(CH3)2 ratio within each of the different microgels (which will determine the surface charge sign and density) as well as the absolute cationic functional group content of the microgel (which strongly influences the average pKa of the PBA functional groups and thus the total number of ionized PBA groups which are able to bind glucose at any given pH). Comparing APBA(+)-22-5 A5 E50 and APBA(+)-17-4 A5 E50, the -COOH/ -N(CH3)2 ratio in the two microgels is almost identical while APBA(+)-22-5 A5 E50 contains slightly more cationic functional groups. Consequently, the change in the surface charge density of APBA(+)-22-5 A5 E50 when glucose is added is slightly higher

- 24 -

SUBSTITUTE SHEET (RUE 26) at pH 7.4 and slightly lower at pH 9 (since a higher percentage of the PBA groups are ionized at pH 7.4 in the microgel containing a higher density of cationic functional groups). However, the gels are similar enough that the observed volumetric swelling behaviours are very similar: -50% volume decreases at pH 7.4 and -70% volume increases at pH 9 are observed in both cases upon the addition of lmg/mL glucose.

In contrast, APBA(+)-27-7 has significantly more cationic functional groups as well as a lower -COOH/-N(CH3)2 ratio when compared to the other two microgels. Thus, the average pK* of the PBA groups in this microgel is significantly lower and more PBA functional groups are ionized at pH 7.4. As a result of this higher degree of PBA ionization, the net cationic charge density observed at Omg/mL glucose is significantly lower and a much higher degree of swelling is observed in the presence of glucose at pH 7.4. Indeed, enough PBA groups are ionized in this case such that the net charge density switches from positive to negative upon glucose exposure, resulting in positive microgel swelling and a two-fold increase in the microgel volume in the presence of lmg/mL glucose.

At pH 9 (which for the APBA(+)-27-7 microgel is significantly above the average pKa of the PBA functional groups), such a high percentage of PBA groups is already ionized prior to glucose exposure that only minimal swelling can be achieved when glucose is added to the suspension. Only a -50% increase in microgel volume is observed when lmg/mL glucose is added. Thus, one can readily control both the absolute and relative magnitudes of the glucose swelling or deswelling responses ofthe amphoteric PBA- containing microgels at different pH values based on both the ratio and total number of cationic and anionic functional monomers used in the microgel recipes.

The observed pH-sensitivity ofthe glucose response can be applied to achieve super- responsive glucose microgel swelling. In the presence of a fixed concentration of lmg/mL glucose, four-fold (APBA(+)-17-4 A50 E50) to ten-fold (APBA(+)-22-5 A5 E50) changes in volume are achieved simply by changing the pH ofthe microgel suspension from 7.4 to 9.0. Thus, the glucose concentration can be used to enhance pH swelling in PBA-microgels. Both the absolute and relative magnitudes of the glucose swelling (or deswelling) responses in amphoteric PBA-microgels can therefore be tuned by controlling the ratio and total number of cationic and anionic functional monomers used to prepare the platform microgels and the amount of PBA grafted to the microgels.

Similar trends are observed when an anionic initiator system is used to prepare the PBA- microgel conjugates. Figure 7 shows the hydrodynamic volume and light scattering intensity responses of an anionically-initiated microgel at pH 7.4 and pH 9.

As more APBA is conjugated, the net surface charge on the microgel at pH 7.4 becomes less negative until eventually switching to positive at high degrees of conjugation, at which point the microgel shrinks instead of swells upon glucose exposure. This negative- to-positive surface charge transition occurs at higher degrees of APBA conjugation than with the cationically-initiated microgels since the residual initiator fragments bound to the microgel are anionic instead of cationic and more -COOH groups must be consumed to allow the cationic charges to control the surface charge. On this basis, the use of anionic initiation is advantageous if one wishes to synthesize microgels which swell in response to glucose since more PBA can be conjugated without inverting the surface charge; conversely, cationically-initiated microgels are better for synthesizing microgels which shrink upon glucose addition since high PBA graft contents will yield more highly charged (and thus more highly responsive) microgels.

Figure 7 also illustrates the dramatic increase in swelling response which can be achieved by lowering the degree of crosslinking in the microgel. The crosslinker feed used to prepare APBA(-)-22-5-2 is approximately 60% lower than that used in the preparation of the cationically-initiated microgels. Even at the intermediate degree of conjugation (the A5 E20 microgel), a three-fold volume increase and an 85% decrease in light scattering intensity is observed upon exposure to lmg/mL glucose. This is a significantly larger response than can be achieved with the cationic microgels even at the highest degree of conjugation tested (see Figure 6). More remarkably, at higher degrees of conjugation (the A5 E50 microgel) a 13-fold volume increase is achieved at pH 9 in the presence of just lmg/mL glucose. This microgel could therefore be applied to the" sensing applications described hereinbefore to achieve even higher resolution for glucose concentration detection.

Based on the results hereinbefore shown, it is expected that similar swelling behaviours could be achieved at physiological temperature and pH 7.4 by conjugating less APBA to the amphoteric microgels to maximize the charge content of the microgels and thereby keep the VPTT above 370C. Table 8 illustrates that physiological pH and temperature activity can be achieved in these amphoteric PBA-containing microgel systems.

Table 8. Hydrodynamic diameter, electrophoretic mobility, and light scattering intensity of APBA- microgels which are responsive at both pH 7.4 and pH 9 and both ambient and physiological temperature. d = hydrodynamic diameter, μe = electrophoretic mobility, φ = light scattering intensity

Microgel Variable pH7.4, 25C pH 7.4, 37C pH9,25C

0mg/mL lmg/mL 0mg/mL lmg/mL Omg/mL lmg/mL Glucose Glucose Glucose Glucose Glucose Glucose

AMPH(+)- d(τιm) 186 3 207 5 141 3 155 3 3153 355 7 22-5 A2 ElO /Z8(XlO-8InWs) -0.90 0.03 -1.17 + 0.05 -0.35 0.01 -0.45 0.01 -1.57 0.03 -1.99 0.05 φ (kcnts/s) 167 2 131 2 3121 294 1 108 1 103 1

AMPH(+)- d(nm) 172 3 190 + 2 144 1 160 3 272 3 3125 22-5A2E15 /4(XKr8ITrVVs) -0.51+0.03 -0.65 + 0.03 -1.26 + 0.02 -1.70 0.05 -1.13 0.03 -1.77 0.07 φ (kcnts/s) 105 1 104 1 148 1 141 1 . 148 1 124 1

AMPH(+)- d(nm) 88 + 1 115 + 5 158 2 186 + 4 152 4 176 4 27-7A5E50 /Z0(XlO"8 πrWs) +0.19 + 0.01 -0.26 0.03 +0.38 0.03 -1.08 0.06 -0.75 0.02 -0.82 0.03 φ (kcnts/s) 227 1 150 2 333 1 203 1 109 1 561

Each of these three microgels exhibits volumetric swelling (between 30-60% increases in volume), decreases in light scattering intensity, and more negative surface charge densities upon exposure to glucose under physiological pH and temperature conditions, exactly as required for the design of a self-regulating insulin release system.

Activity at physiological salt conditions is rarely achieved for hydrogels whose stimulus- responsive swelling is driven by electrostatics. Charge screening minimizes the impact of any net change in polymer charge density under such high salt concentrations. However, the observed anti-polyelectrolyte behaviour of amphoteric microgels facilitates swelling responses at high ionic strengths in these novel microgels. Table 9 shows the behaviour of APBA(+)-22-5 A5 E50 microgel at ambient temperature. Table 9. Diameter, volume, and light scattering intensity response of the APBA(+)-22-5 A5 E50 microgel at various salt and glucose concentrations (pH 7.4, 250C). The volume ratio and relative light scattering intensity are normalized based on the Omg/mL glucose and 0.15M salt concentration result.

Glucose Salt Diameter Volume Light Scattering Intensity

Concentration Concentration (nm) Ratio Absolute Relative

(mg/mL) (mol/L) (kcnts/s)

0 0 163 2 1.22 + 0.06 70 + 1 0.57 + 0.01

0 0.15 153 1 1 121 +1 1

1 0.15 171 2 1.37 + 0.06 113 + 1 0.93 + 0.02 5 015 197 + 5 2.13 + 0.07 99 + 3 0.81 + 0.04

The PBA-microgel conjugate behaves as a typical amphoteric polymer in that only a rninimal decrease is observed in the microgel diameter upon the addition of 0.15M NaCl. Indeed, the amphoteric microgels exhibit more narrowly dispersed particle sizes and enhanced colloidal stability in the presence of higher salt concentrations compared to the low ionic strength buffer. The addition of glucose induces a significant, systematic swelling response, with a -40% volume increase occurring at lmg/mL glucose and a 110% volume increase occurring when 5mg/mL glucose is added. Consequently, this microgel system could be used "as is" in a laboratory environment to detect glucose concentrations in physiological fluids at ambient temperature, exploiting either or both of the particle size and light scattering responses to glucose.

Anionic-initiated microgels exhibit the same glucose-sensitive swelling responses at high ionic strengths while providing improved colloidal stability at physiological temperature to achieve "true" physiological swelling activity, as shown in Table 10.

Table 10. Ratio changes in hydrodynamic volume and light scattering intensity for anionically-initiated amphoteric microgels exposed to lmg/mL glucose (pH 7.4, 0.15M ionic strength)

Microgel Temperature Ratio (*(lmg/mL GIucose)/x(0mg/mL Glucose)

(0C) x = hydrodynamic jc = light scattering volume intensity

25 1.10 + 0.03 0.85 0.01

APBA(-)-17-5-2 A5 E50 37 1.36 + 0.02 0.84 0.01 25 1.60 0.05 0.81 0.01 APBA(-)-17-5-2 A5 E20 37 1.76 + 0.07 0.88 0.02

Like the cationic-initiated microgels in Table 9, the anionic-initiated microgels shown in

Table 10 contract by ~10% upon the addition of 0.15M NaCl but remain colloidally stable under physiological conditions. The anionic initiator system used to prepare these microgels is more effective at both stabilizing the microgel under the high ionic strength conditions and shifting the VPTT of the microgel above physiological temperature, preventing the aggregation observed with the cationic-initiated microgels at 37C. Furthermore, both microgels swell more in response to lmg/mL glucose at 370C than at 25C and the A5 E20 microgel swells more in response to glucose despite containing fewer PBA groups. Both of these observations are indicative of thermal amplification of the glucose-driven phase transition. Macroscopic microgel deswelling is observed at ~35C for the A5 E50 microgel and ~36C for the A5 E20 microgel, both just below the test temperature of 37C. Thus, the VPTT shift observed upon glucose binding can be still applied at high ionic strengths to drive optimized changes in microgel diameter as the glucose concentration increases under high ionic strength conditions.

Furthermore, in parallel to the anionic microgels in Figures 2 and 3, amphoteric microgel swelling can be both amplified and suppressed by the thermal volume phase transition. Table 11 shows the volumetric swelling and light scattering intensity responses of amphoteric PBA-microgels containing increasing PBA contents.

Table 11. Volumetric and light scattering swelling ratios for anionic-initiated PBA-microgel conjugates with increasing PBA contents in the presence of lmg/mL glucose (pH 7.4, 37C, 0.15M ionic strength)

Microgel Ratio jc(lmg/mL Glucose)/jc(0mg/mL Glucose) x = Volume x = Scattering Intensity

AMPH(-)-23-6-2 A2 E20 1.32 + 0.09 0.92 0.01 AMPH(-)-23-6-2 A2 E50 1.66 + 0.07 0.91 + 0.01 AMPH(-)-23-6-2 A2 ElOO 1.07 + 0.02 0.98 + 0.01

Glucose-responsive swelling increases with the PBA content of the microgel until the 37C test temperature is higher than the microgel VPTT both in the presence and absence of glucose. On this basis, glucose swelling in the A2 E20 microgel is driven primarily by ionic interactions, glucose swelling in the A2 E50 microgel is enhanced by the thermal phase transition, and glucose swelling in the A2 ElOO microgel is suppressed by the thermal phase transition. Again, this response directly parallels the anionic microgel responses at low salt described hereinbefore. This observed thermal phase transition enhancement of the glucose-driven phase transition is particularly useful in the high ionic strength physiological environment. The ionization-based mechanism of glucose swelling is suppressed at high salt conditions via charge screening while the VPTT shift, although somewhat affected, is significantly less sensitive to the presence of salt Consequently, this microgel system exhibits both good colloidal stability and significant glucose swelling capability (i.e. up to two-fold volumetric increases) at physiological pH, temperature, and ionic strength.

Both swelling and shrinking responses can also be observed in the amphoteric microgels under physiological conditions, as illustrated in Table 12.

Table 12. Volumetric swelling ratio for anionic-initiated PBA-microgel conjugates in the presence of different glucose concentrations (pH 7.4, 37C, 0.15M ionic strength). The electrophoretic mobilities are measured in 5OmM buffers to properly resolve the microgel charge

Microgel Electrophoretic V(Jt mg/mL Glucose)/V(0mg/mL Glucose)

Mobility x = lmg/mL Λ: = 2mg/mL (xlO 8 mz/Vs)

AMPH(-)-28-8-2 A2 E20 -2.53 0.03 1.27 + 0.03 1.38 + 0.02 AMPH(-)-28-8-2 A2 E50 -0.60 0.04 0.98 + 0.02 1.01 + 0.01 AMPH(-)-28-8-2 A2 E100 +0.59 + 0.06 0.91 + 0.01 0.83 + 0.01 An amphoteric microgel with a net negative charge in the absence of glucose (the A2 E20 microgel) swells systematically when exposed to 1 mg/mL glucose while a microgel with a net positive charge prior to glucose addition (the A2 ElOO microgel) deswells systematically with the glucose concentration. Hence, all the microgel responses observed in amphoteric microgels at low ionic strengths can be replicated at high ionic strengths. Increasing the magnitude of the glucose responses observed at high ionic strength can likely be accomplished by increasing the functional group loading of the platform microgel to allow more of the glucose-responsive PBA groups to be grafted without lowering the VPTT of the microgel below physiological temperature.

Cell Flocculation Using PBA-Functionalized Amphoteric Microgels

Cell surfaces carry often quite thick layers of carbohydrates, some of which contain the cis-άiol stereochemistry required for PBA binding. PBA-functionalized microgels can induce the flocculation of both E. coll and P. pastoris cells under a range of solution conditions. However, the particular biotechnology usefulness of this strategy lies m the use of glucose to regulate the flocculation of cells via two possible mechanisms. When amphoteric PBA-microgels are used as the flocculant, cell flocculation can be induced electrostatically via inversion of the microgel surface charge as the glucose concentration increases. Figure 8 shows the particle size distributions of E. coli. suspensions when two different PBA-functionalized amphoteric microgels are added in the presence of different glucose concentrations. Table 13 gives the corresponding electrophoretic mobilities for each of the flocculating microgels and the E. coli. bacteria as a function of glucose concentration. Flocculation is indicated by the appearance of large particles in the particle size distributions.

Table 13. Electrophoretic mobilities of amphoteric PBA-microgels used in flocculation testing over a range of glucose concentrations (pH 9, 5OmM, carbonate buffer)

Microgel Electrophoretic Mobility (x 10 s m2/Vs)

Omg/mL lmg/mL 10mg/mL 20mg/mL Glucose Glucose Glucose Glucose

AMPH(+)-17-5 A5 E20 -0.18 + 0.12 -0.24 0.09 -0.35 0.05 -0.65 + 0.03 AMPH(+)-17-5 A5 E50 +0.25 0.01 +0.14 0.05 -0.21 0.05 -0.46 0.03 APBA(+)-17-5 A5 E100 +0.43 0.06 +0.21 0.07 -0.05 0.10 -0.38 0.07 E. coli. bacteria -2.09 + 0.07 N/A N/A N/A

In both Figure 8(a) and Figure 8(b), E. coli. flocculation is observed at low glucose concentrations while no significant flocculation occurs at high glucose concentrations. Correspondingly, the surface charge on the microgel switches from cationic at low glucose levels to anionic at higher glucose levels as the PBA equilibrium shifts and more anionic PBA sites are generated. Since E. coli. cells are highly anionic at the physiological test pH, the microgels and cells would electrosterically repel each other at high glucose concentrations and electrostatically attract each other at low glucose concentrations. This charge inversion mechanism permits on-off switching of cell flocculation as a function of glucose concentration. Without intending to be bound by theory, the probable mechanism of glucose-controlled cell flocculation of E. coli. is illustrated in Scheme 1.

- 31 -

SUBSTiTUTE SHEET W HE 2W Increase [Glucose]

Electrostatic Attraction Electrostatic Repulsion

Scheme 1. Mechanism of glucose-regulated charge-controlled cell flocculation The glucose concentration at which this switching occurs can be controlled by the PBA content of the microgel. For the AMPH(+)-17-5 A5 E50 microgel (Figure 8(a)), fewer PBA groups are conjugated to the microgel and lOmg/mL glucose is sufficient to switch the surface charge of the microgel from positive to negative and "turn off the flocculation. However, in AMPH(+)-17-5 A5 ElOO (Figure 8(b)), more -COO" groups are consumed by PBA conjugation and more PBA groups must be ionized to switch the microgel surface charge from positive to negative. Thus, the cells flocculate in the presence of lOmg/mL glucose and 20mg/mL glucose is required to block flocculation. The PBA content of the microgel can be used to control the interactions between the microgels and the cells in the absence of glucose. Figure 9 shows the particle size distributions of E. coli. cells in the presence of microgels containing varying PBA contents.

Both the A5 E50 and A5 ElOO conjugated microgels are cationic in the absence of added glucose (Table 13) such that E. coli. flocculation is induced via the electrostatic mechanism of Scheme 1. However, the A5 E20 microgel carries a net anionic charge in the absence of glucose and, correspondingly, does not induce cell flocculation. Indeed, the particle size profile given in Figure 9 shows two distinct peaks corresponding to both the unflocculated E. coli. cells (centred at ~lμm) and the free microgel particles (centred at ~0.15μm). This suggests that like surface charges on the microgel and the cell not only block cell flocculation but also switch off all significant interactions between the microgels and the cell surface. This observation can be exploited to recover harvested cells after flocculation by switching off the microgel-cell adhesion. Initial experiments suggest that up to 30% cell recovery can be achieved by adding lOmg/mL glucose to the AMPH(+)-17-5 A5 E50 microgel-flocculated cell suspension, switching the surface charge on the microgel to negative and "turning off the cell-microgel interactions. While 100% recovery is likely unrealistic given the additional interactions which may occur when electrostatic interactions bring the microgels and cells in close proximity (i.e. hydrophobic interactions between non-ionized PBA groups and hydrophobic sites on the cell surface), the microgel surface composition can be optimized to maximize the degree of cell recovery possible upon charge inversion.

This glucose-responsive charge inversion flocculation strategy could be applied for the selective harvesting of cells at specific phases in their growth cycles. The glucose concentration added during cell culturing and/or the PBA content of the flocculating amphoteric microgel can be tuned to flocculate the growing cells at a selected interval in their growth cycle.

P. pastoris cells also carry a high anionic charge (μe = -1.75 x 10"8 mVVs) and can be flocculated via the same charge inversion mechanism used to flocculate E. coli. However, P. pastoris cells can also be flocculated by a specific PBA-cell surface carbohydrate binding mechanism. Figure 10 reports the particle size distributions for the P. pastoris suspensions before and after microgel addition in the presence of glucose while Table 14 shows the corresponding relative turbidities of the cell suspensions and Table 15 reports the relevant electrophoretic mobilities of both the cells and the microgels under the flocculation conditions.

Table 14. Relative turbidity results for the flocculation of P. pastoris with PBA-modified microgels (pH 9, 5OmM carbonate buffer, 25C). A 1 hour wait time was used between microgel addition and sampling. Relative turbidity values were calculated by dividing the turbidity of the flocculated mixture by the sum of the unflocculated yeast, microgel, and buffer turbidities.

Microgel 0mg/mL Glucose 2mg/mL Glucose lOmg/mL Glucose

Turbidity Relative Turbidity Relative Turbidity Relative (NTU) Turbidity (NTU) Turbidity (NTU) Turbidity

APBA-22-0 A2 E20 19.3 0.63 N/A N/A APBA-22-0 A5 E20 16.4 0.53 31.7 1.03 30.6 1.00 APBA-22-0 A5 ElOO 13.6 0.44 23.5 0.77 29.7 0.97 Table 15. Electrophoretic mobilities of PBA-tnodified microgels and yeast used in the flocculation experiments under different glucose concentrations (pH 9, 5OmM carbonate buffer, 25C)

Microgel Electrophoretic Mobility (x IQ-" mz/V s)

Onig/mL Glucose 2mg/mL Glucose lOmg/mL Glucose

APBA-22-0 A2 ElO -0.88 + 0.11 N/A N/A

APBA-22-0 A2 E20 -0.68 + 0.08 N/A N/A

APBA-22-0 A5 E20 -0.54 0.01 -1.10 0.03 -1.25 + 0.04

APBA-22-0 A5 ElOO -0.62 + 0.07 -1.27 0.08 -1.36 +0.09

P. pastoris -1.75 + 0.07 -1.60 0.05 -1.78 + 0.07

Table 15 indicates that both the microgels and the cells carry significant anionic charges at all tested glucose concentrations at pH 9, the pH value used for the P. pastoris flocculation experiments. Consequently, based on the lack of interaction observed in Figure 9 between microgels and cells carrying like charges, neither cell flocculation nor microgel-cell interactions would be anticipated under these conditions. However, floes become visible in the particle size distributions (Figure 10) and tlhe relative turbidities the cell suspensions significantly decrease (Table 14) as PBA-microgels are added to P. pastoris suspensions. These observations suggest that the PBA-functionalized microgels can mediate P. pastoris flocculation in the absence of glucose.

Furthermore, cell flocculation can be turned off as glucose is added to the system despite the absence of a charge inversion event which can alter the electrostatic interactions between the microgels and the yeast cells. Without intending to be bound by theory, this observations can be rationalized based on specific binding interactions between cell surface carbohydrates and the PBA microgels. The probable mechanism is illustrated in Scheme 2.

Inhibits

Scheme 2. Mechanism of glucose-regulated PBA-controlled cell flocculation

In the presence of two different c/s-diol containing carbohydrates, PBA groups will interact with both carbohydrates according to a competitive binding mechanism. The

- 34 -

SUBSTiTUTE SHEET {RULE 26) ratio of each carbohydrate in solution and the affinity of each carbohydrate for PBA binding (as indicated by the equilibrium binding constant K0) would control the relative amounts of each carbohydrate bound to a PBA-functionalized microgel. In the absence of glucose, cell surface carbohydrates are the only cis-diol carbohydrate in the suspension and can therefore bind effectively with PBA sites on the microgels, stimulating the observed cell flocculation. As more glucose is added, glucose competitively binds with the PBA groups, blocking these sites from interacting with cell surface carbohydrates. Thus, the degree of flocculation achieved upon microgel addition decreases systematically with the glucose concentration until cell flocculation is essentially "turned off' at high glucose loadings. Indeed, at lOmg/mL glucose, no significant aggregation is detected in the particle size distributions in Figure 10 and no decrease. in turbidity is noted upon microgel addition in Table 14.

The difference in the flocculation behaviors observed in the A5 E20 and A5 ElOO microgels in Figure 10 can also be understood based on the competitive binding mechanism. At lower PBA loadings (the A5 E20 microgel), less glucose is required to effectively block all of the available PBA sites on the microgel. Correspondingly, cell flocculation is significantly limited at 2mg/mL glucose and entirely turned off at lOmg/mL glucose. In the more highly functionalized A5 ElOO microgel, more PBA groups are present and more glucose is required to competitively inhibit PBA-cell surface carbohydrate interactions. Thus, more and larger floes form (Figure 10) and the relative turbidity decreases more (Table 14) in the presence of 2mg/mL glucose when the A5 ElOO microgel is used as the microgel fiocculant.

The degree of cell flocculation observed upon microgel addition can also be tuned according to the overall PBA graft content in the microgel, as shown in Figure 11.

No glucose is added in any of these experiments such that the total cell flocculation capacity of each of these microgels can be compared directly in the absence of a competitive inhibitor. As the PBA graft content of the microgel increases, more binding sites become available for cell surface carbohydrates and more (and larger) cell aggregates are formed. In parallel, Table 14 indicates that the relative turbidity of the cell

- 35 -

SUBSTiTUTE SHEET (RtIE 26) suspension decreases systematically as the PBA content of the microgel used for the P. pastoris flocculation increases. Both of these observations provide additional evidence for the specific PBA-cell surface carbohydrate flocculation mechanism given in Scheme 2. Furthermore, initial screening indicated that 20-40% of cells can be released from the cell-microgel aggregates via competitive binding upon the addition of glucose, again suggesting the potential of these PBA-microgels as recyclable cell recovery vehicles.

The selective removal of one cell from a mixture of different cell types is possible given the specific PBA-cell surface carbohydrate mechanism of cell flocculation. Figure 12 shows how the APBA-22-0 A5 El 00 microgel can flocculate P. pastoris cells but cannot flocculate E. coll cells in a pH 9 salt solution.

Over 95% of the P. pastoris cells in suspension are flocculated by the APBA-22-0 A5 ElOO microgel and no free microgel peak is observed in the particle size distributions just 10 minutes after microgel addition. In contrast, no E. coll flocculation is achieved upon microgel addition and microgels do not appear to adhere to the E. coll cells given the large residual microgel peak observed in the particle size distribution profile. This observation stands even when the total microgel loading to the flocculation system is doubled or the wait time after gel addition is increased from 10 minutes to 2 hours. Thus, P. pastoris can be flocculated selectively in the presence of E. coll via specific PBA-cell surface interactions.

While cell surface carbohydrate compositions are extremely difficult to specify, general trends are observed which can rationalize the different selectivities observed in the PBA- based flocculation of E. coli. and P. pastoris. The outer membrane of E. coll is comprised primarily of peptidoglycan, the two major components of which (N- acetylglucosamine and N-acetylmuramic acid) do not contain any cis-ά\o\ functional groups. A variety of lipopolysaccharides (some of which may contain cis-diol groups) also protrude from this surface into the solution. Conversely, in yeast cells like P. pastoris, the cell wall is comprised primarily (>80%) of glucans and mannans, both of which contain cώ-diol groups. Furthermore, the glycocalyx protruding from the cell surface is rich in glycolipids and glycoproteins which contain cis-diol stereochemistry.

- 36 -

SUBSTITUTΕ SHEET (HULE 26) Thus, the surface of a P. pastoris cell would present significantly more cw-diol groups to the microgel than would the E. coli. surface. This compositional difference may explain the higher observed flocculation in the P. pastoris yeast system.

Based on these results, it should be possible not only to flocculate biological agents using the PBA-carbohydrate interaction as the primary driver of the flocculation but also to selectively flocculate specific types of cells in the presence of other cells according to the surface chemistry of the cells (specifically, the net charge and types and quantities of carbohydrates expressed on the surface of the cell). This would be useful in bioseparation protocols for isolating particular cell types from complex mixtures or selectively removing cells with particular surface chemistries from inhomogeneous cell cultures.

A variety of physical and chemical cues can be used to alter the speed of the flocculation process and the size of the floes. Commercial Type II Saccharomyces cerivisae cells were used as a model system to evaluate the effect of the carbohydrate chemistry, ionic strength, and pH on the flocculation process. The effect of the carbohydrate of the inhibition of cell flocculation is shown in Figure 13, in which glucose is replaced by sucrose as the carbohydrate inhibitor.

:The association equilibrium constant K0 for phenylboronic acid-carbohydrate binding is significantly higher for glucose (K0 = 4.6) than for sucrose (K0 = 0.7), owing to the lower number of cis-diol groups in sucrose per unit mass and the steric inhibitions to PBA polymer complexation presented by the bulkier disaccharide structure. Thus, cell surface carbohydrates should more successfully compete with added sucrose to promote cell flocculation compared to glucose. Indeed, Figure 13 indicates that significant flocculation occurs in the presence of 20mg/mL sucrose while no effective flocculation is observed in the presence of 20mg/mL glucose. However, sucrose does significantly slow the flocculation rate and reduces the size of the resulting floes compared to the case in which no carbohydrate inhibitor is added. Thus, the higher the PBA-carbohydrate binding constant, the more successfully the carbohydrate competes with cell surface carbohydrates to inhibit flocculation and the slower the flocculation rate.

The effect of electrolyte on PBA-specific flocculation is shown in Figure 14.

As the salt concentration increases, the electrostatic repulsion between the highly anionic yeast cells and the anionic microgel is screened. Thus, even when glucose can successfully block the PBA binding sites in the microgel (as is the case at 20mg/mL glucose), simple diffusion can orient the microgels and cells in close enough proximity that non-specific interactions such as hydrophobic binding can induce microgel-cell flocculation. The relative colloidal instability of the microgels in the 50OmM total ionic strength experiments also promotes both microgel aggregation and microgel precipitation on to the cell surface. Thus, although the cells are themselves stable over at least several hours at the high ionic strength conditions, microgel precipitation provides an anchoring site on the cell surface for the formation of cell aggregates. Similar enhanced cell flocculation at high ionic strengths is observed when sucrose or cellobiose is used as the carbohydrate inhibitor, both of which have weaker binding PBA binding constants than glucose.

The same non-specific flocculation effect is observed at pH 4, at which point the microgel has a relatively weak anionic charge and no specific PBA-carbohydrate interactions are present. Figure 15 shows the effect of pH on cell-microgel flocculation.

At both pH 9 and pH 10, the PBA groups are ionized in their tetrahedral, glucose-binding form and both the microgel and the yeast carry large anionic charges. Consequently, electrostatic repulsion keeps the microgel and yeast apart to minimize non-specific adhesion while glucose can competitively inhibit the specific PBA-cell surface carbohydrate interactions driving the yeast flocculation. Indeed, while <1% flocculation is observed at both pH 9 and pH 10 in the presence of 20mg/mL glucose, nearly 100% flocculation efficiency is observed at both pH 9 and pH 10 if glucose is not added to block the specific PBA-cell binding. However, at pH 4, the surface charges on both the yeast and the microgel are significantly lower and some non-specific flocculation is observed. The PBA residues cannot bind to cell surface carbohydrates at this acidic pH; correspondingly, the same degree of flocculation is observed whether or not glucose is added to the suspension. These results suggest that both high net surface charges on the microgel and the yeast (rninirnizing non-specific interactions) and glucose inhibition of PBA-cell surface carbohydrate binding (minimizing specific interactions) are required to effectively turn the flocculation process "on" and "off' via external intervention.

Based on these results, the selectivity of the PBA-cell interactions can effectively be tuned by the external salt concentration, the pH of the suspension, and the type of inhibiting carbohydrate added. This multi-variable control over cell flocculation (or, more generally, cell adhesion) may be useful for controlling the flocculation rate, the critical glucose concentration for turning the flocculation "on" and "off', and the size the resulting cell aggregates.

Applications of Cell-Selective Adhesion

The concept of selective PBA-mediated flocculation may be extended to a number of other applications. Cells grown under different medium conditions may be differentiated since the nature of the growth medium can influence the carbohydrate composition at the surface of cells. Biological signals may be mediated via specific cell-microgel interactions given the importance of cell surface carbohydrates in regulating cell signaling events. Different types of carbohydrates can also be targeted by changing the chemical structure of the PBA binding group used. As an example, using a cyclic boronic acid increases the cone angle in the boronic acid residue to enhance the selectivity of boronic acid binding to glycoside carbohydrates over other sugars. Using these strategies, mammalian cells expressing specific carbohydrate markers may be targeted using boronic acid-functionalized microgels, a useful technology for both cell recovery and cell signaling. A variety of fluorescent PBA residues have also been designed which would facilitate the selective imaging of specific cell types in complex mixtures.

- 39 -

SUBSTiTUTC SHEET (RULE 26) Targeted drug delivery should also be possible by tuning the PBA-microgels to bind to cells displaying specific carbohydrate markers. The PBA functionality of the microgel can anchor the delivery vehicle to a targeted cell surface while the gel phase can control the release rate of a specific drug at the target site. PBA binding to a cis-diol surface would also shift the PBA equilibrium to generate more anionic PBA sites, enabling control over drug release via both the diffusion and charge-controlled mechanisms described in Scheme 1. Two examples are of particular interest in this context. A typical marker of cancer cells is the over-expression of sialic acid, a cis-diol containing carbohydrate with a high affinity binding with PBA. Similarly, the high sialic acid content in mucous membranes may be applied to use PBA-microgels as drug delivery vehicles for the oral or nasal delivery of pharmaceutical vehicles.

Glucose Sensors

The use of PBA-microgels as glucose sensors is contemplated, flowing from the systematic diameter and light scattering changes observed as the glucose concentration is increased in both low salt (Figure 3) and high salt (Table 12) conditions, hi particular, thermally-enhanced glucose swelling observed within specific glucose concentration ranges in Figure 3 is beneficial to produce sensors with enhanced glucose responses over targeted glucose concentration ranges. Figure 16 shows that linear diameter and light scattering intensity responses can be achieved with the microgel sensors over specific ranges of glucose concentration by controlling the PBA content of the microgel.

As the PBA content of the microgel increases, the magnitude of the overall glucose response increases and the glucose concentration range over which linear particle size and light scattering intensity responses are observed narrows. These trends can be rationalized based on the PBA equilibrium shown in Scheme 3.

Scheme 3. Reversible binding of glucose to (alkylamido)phenylboronic acid

- 40 -

SUBSTiTUTE SHEET (RULE 26} As the number of PBA groups present in the microgel matrix increases, a smaller change in the glucose concentration can drive larger increases in the gel-phase degree of ionization. However, at high PBA conjugation yields, the PBA functional groups become clustered along the polymer chain and charge-driven swelling frustration is observed. The more PBA groups are grafted to the microgel, the more clustered PBA "blocks" are present such that effective gel swelling occurs primarily at lower degrees of ionization. This effectively reduces the glucose concentration range over which significant swelling responses are observed. Thus, according to the number and distribution of PBA groups in the microgel, glucose sensors with large responses over specific glucose concentration ranges can be synthesized.

The particular sensors shown in Figure 16 are well suited for laboratory measurements of glucose given their responsiveness at room temperature compared to physiological temperature. These microgels could also be applied in the construction of a microgel sensor device comprised of a three-dimensional colloidal array of the highly monodisperse PBA-microgels constructed using well-defined strategies. Glucose concentration changes could then be tracked both by changes in the overall turbidity of the microgel film and the color of light refracted from the highly ordered microgel arrays as the inter-lattice spacing changes with the degree of swelling in the individual glucose- responsive microgels.

The observed adjustability of the glucose concentration response range based on the PBA content of the microgel further suggests the potential of tuning the glucose sensors for different glucose-containing environments. For example, the glucose concentration in the intracellular space is slightly lower than that of the arterial blood (normally ~3.9-4.8mM as opposed to 5.2-7.2mM) while the normal glucose concentration in tears is 5-10-fold lower than that in blood. Both concentrations are directly related to the blood glucose level and could be used to indirectly measure blood glucose and/or (particularly in the intercellular region) regulate drug release. The pH of 7.2-7.3 in these environments is still sufficiently high to generate glucose responses in the amphoteric microgels. Thus, the PBA-microgels should be useful for both laboratory and in vivo sensing in a range of different environments.

Insulin Uptake and Release The phenylboronic acid-grafted microgel technology should also be useful in the self- regulation of blood glucose levels within safe ranges over a variety of time frames. Scheme 4 shows two potential mechanisms of feedback-controlled insulin release facilitated by the PBA-microgel systems.

Scheme 4. Mechanisms of glucose-responsive insulin release in microgels: (a) diffusion release mechanism; (b) charge inversion release mechanism. The dark triangles represent insulin molecules. The diffusion release mechanism (Scheme 4(a)) relies on the macroscopic swelling observed in PBA-microgel networks as the glucose concentration increases. At low blood-glucose levels, little glucose-induced swelling is observed in the microgel and the small pore size of the microgel network resists insulin diffusion out of the gel network. As the blood glucose level increases, more glucose will bind to the microgel, shifting the PBA ionization equilibrium to generate more charged PBA groups. This causes the microgel to swell, facilitating the release of insulin. As the released insulin acts physiologically, the blood glucose level will drop. Since the by-product (glycogen) does not contain any of the czs-diol groups which bind to PBA, the equilibrium will shift back toward the uncharged PBA species and the gel will deswell, reducing the diffusion rate of insulin out of the microgel and avoiding "overshoots" in blood glucose correction.

Alternately, the charge inversion mechanism (Scheme 4(b)) relies on the switching of the net charge density in the microgel upon glucose binding. Insulin is a protein of mass 5.5kDa and isoelectric point of 5.4 and thus carries a negative charge at physiological pH. Therefore, microgels which carry a net cationic charge in the absence of glucose at physiological pH can bind insulin via electrostatic attraction. As the glucose concentration increases, the PBA equilibrium shifts to generate more anionic charges and switch the net charge on the microgel to anionic, driving insulin release via electrostatic repulsion. Both mechanisms have a feedback mechanism (microgel deswelling in (a) or anionic charge density reduction in (b)) which reduce the rate of insulin release as the insulin acts and the glucose concentration in the environment is reduced.

Preliminary experiments have been performed to confirm the potential of amphoteric PBA-microgels in this application. The AMPH(-)-28-8 A5 E50 microgel was selected as a model system given its high PBA content (-70% of -COOH groups converted to PBA groups) and its high amine content (maximizing the potential for charge inversion and lowering thepKa of the maximum number of PBA groups). The AMPH(-)-28-8 A2 E20 microgel which contains the same base functional group composition but a significant lower PBA content (-25% -COOH conversion) is also evaluated to test the effect of the PBA content on the insulin responses. Table 16 shows the uptake and Table 17 shows the release of insulin achieved using this microgel as a function of both pH and glucose concentration and the corresponding surface charges on the microgel.

Table 16. Insulin uptake (mg/mg dry microgel) for amphoteric PBA-microgels of different PBA contents in pH 4 citrate buffer and pH 7.4 phosphate buffer (1OmM ionic strength).

Microgel Electrophoretic Mobility Insulin Uptake (xlθ~8 mVVs) (mg insulin/mg microgel) pH 4, 10mM pH 7.4, IQmM pH 4, 10mM pH 7.4, 1OmM

AMPH(-)-28-8 A2 E20 -1.86 0.04 -1.71 0.02 6.18 0.05 0.26 + 0.12

AMPH(-)-28-8 A5 E50 -1.03 0.04 +0.30 + 0.02 3.33 0.08 2.64 + 0.15

Table 17. Insulin release (% of insulin in each of the complexes described in Table-9.1) for amphoteric PBA-microgels of different PBA contents over 24 hours. Insulin loading was conducted using a pH 4, 1OmM citric acid buffer. Release experiments are performed under physiological conditions (pH 7.4, 37C, 15OmM ionic strength) while the mobility values are measured in a 1OmM pH 7.4 phosphate buffer.

Microgel Electrophoretic Mobility of Insulin Release Insulin-Microgel Complex (% of insulin in microgel) (xlO 8 mVVs)

0mg/mL 2mg/mL 0mg/mL 2mg/mL Glucose Glucose Glucose Glucose

AMPH(-)-28-8 A2 E20 -1.36 0.04 -1.45 0.04 23.4 0.7 ' 24.2 1.4 AMPH(-)-28-8 A5 E50 -0.54 0.03 -0.60 0.04 39.3 1.3 46.4 3.2

- 43 -

SUBSTITUTE SHEET (RUtE 26) Insulin uptake appears to be driven primarily by electrostatic interactions between the amphoteric microgels and the insulin protein. At pH 4, insulin carries a net cationic charge while the microgels are anionic due to peptization by the divalent citric acid buffer salt. Consequently, extremely large insulin uptakes are observed in both microgels. At pH 7.4, insulin carries a net anionic charge while the A2 E20 microgel is anionic and the A5 E50 microgel is cationic. Correspondingly, minimal insulin uptake is achieved with the A2 E20 microgel while a large insulin uptake is facilitated by the A5 E50 microgel. Thus, the amount of insulin uptake is a function of the electrostatic interactions between, the amphoteric microgel and insulin at the pH value used for protein uptake. When the pH 4-loaded insulin-microgel complexes are redispersed under physiological conditions (pH 7.4, 15OmM ionic strength, 37C), insulin is released from both microgels. However, the A2 E20 microgel (low PBA content) shows no glucose- specific insulin release while the A5 E50 microgel (high PBA content) selectively releases more insulin in the presence of higher glucose levels. This observation can be understood in the context of the charge inversion mechanism for insulin release (Scheme 4(b)) based on the microgel electrophoretic mobility data given in Figure 17.

APBA(-)-28-8 A2 E20 carries a net anionic charge at physiological pH regardless of the glucose concentration such that microgel-insulin repulsion occurs whether or not glucose is present and no significant difference is observed in the insulin release when glucose is added. Conversely, at pH 7.4, APBA(-)-28-8 A5 E50 carries a slight net cationic charge in the absence of glucose (pl~ 7.8) and a slight net anionic charge in the presence of glucose (pl~ 7.3). Therefore, anionic insulin is attracted to the microgel in the absence of glucose but repelled from the microgel when glucose is added, resulting in the increased observed insulin release at 2mg/mL glucose. It is also important to note that the net charge of each of the insulin-microgel complexes is negative regardless of the glucose concentration (Table 17). Thus, the protein aggregation and immune responses observed when cationic polymers are present in the in vivo environment can be avoided even under conditions at which the bare microgel would carry a net cationic charge. Based on these preliminary observations, amphoteric PBA-microgels should be useful for facilitating glucose-responsive insulin release. To do so, one would need to improve the microgel pi shift observed upon glucose binding. This may be achieved by maximizing the PBA content of the microgel and, correspondingly, the functional group content of the base amphoteric microgel platform. Diffusion-controlled release could also be promoted by controlling the crosslinker density and distribution such that the change in the average pore size as glucose is added significantly enhances the diffusion of the 5.5kDa insulin molecule out of the microgel matrix. Finally, the crosslink density, particle size, and charge ratio in the microgel must be balanced to facilitate physiologically-appropriate release rates in vivo. The formation of microaggregates may be desirable in this context to increase the overall size of the microgel and thus control insulin release over a longer period of time.

The other key observation from these preliminary studies is that either or both of the cationicranionic charge density and the pKa of the functional groups in the amphoteric microgel can be adjusted to facilitate electrostatic uptake and release. Ideally, an amphoteric microgel would be designed to mirror the charge density of a target molecule under the uptake conditions and parallel the charge density of a target molecule under the optimal release conditions. Thus, electrostatic attractions between the microgel and the target are maximized under the uptake conditions while electrostatic repulsion between the microgel and the target are maximized under the release conditions. This ability to tune the charge profile of the microgel to match that of a target molecule and/or target cell may be applied to extend amphoteric microgels as responsive binding agents for other proteins, pathogens or biomolecules with complex charge profiles. CONCLUSIONS

(1) Phenylbdronic acid-modified poly(N-isopropylacrylamide)-based microgels can be designed to exhibit reversible swelling responses to changes in the glucose concentration. (2) The underlying functional group distribution controls the relative changes in the turbidity and particle volume as glucose is added while the total number of PBA groups present in the microgel controls the absolute magnitude of the responses. (3) The thermal phase transition can be applied to both amplify and suppress ionization- driven glucose swelling. Both the temperature and glucose concentration ranges over which glucose phase transition amplification and suppression are achieved can be tuned by changing the underlying composition of the microgel.

- 45 -

SUBSTiTUTE SHEET (RULE 2S) (4) Amphoteric PBA-functionalized microgels can be designed which can either shrink or swell in response to increases in the glucose concentration according to the solution pH and the cationic/anionic charge ratio in the microgel network. (5) Amphoteric microgels are colloidally stable under physiological conditions and can exhibit up to two-fold increases in volume upon exposure to lmg/mL glucose at pH 7.4, 37C, and 0.15M ionic strength. This is believed to be the first reported microgel system to achieve such physiological responses.

(6) Glucose sensors can be designed based on PBA-functionalized microgels which exhibit linear particle size and scattering intensity responses over specific glucose concentration ranges. These linearly-responsive glucose concentration ranges are tunable according to the PBA graft content in the microgel.

(7) Amphoteric microgels can be designed to bind large amounts of insulin (-3-6 times the mass of the dry microgel in the suspension) by matching the charge profiles of the insulin and the amphoteric gel network at the uptake pH.

(8) Glucose-sensitive insulin release should be possible by inverting the microgel charge Upon glucose exposure in physiological conditions.

(9) Glucose can be used to selectively inhibit the flocculation of biological cells by PBA- functionalized microgels by inverting the electrostatic charge on the microgel surface and/or the competitively binding to PBA sites in the microgel to block the specific interaction between PBA groups and cis-dio\ carbohydrates expressed. on cell surfaces.

(10) P. pastoris cells can be selectively flocculated in the presence of E. coli. cells by exploiting differences in the cell surface carbohydrate compositions.

(11) The ionic strength, pH, and carbohydrate inhibitor selected for the flocculation can be applied to tune the aggregation rate and the floe size of PBA-microgel/cell aggregates.

- 46 -

SUBSTITUTE SHEET (PAiLE 26)

Patent Citations
Cited PatentFiling datePublication dateApplicantTitle
CA2044021C *7 Jun 19914 Jun 2002Roderick Glyn RylesCross-linked anionic and amphoteric polymeric microparticles
CA2408312A1 *23 May 200120 Dec 2001Ondeo Nalco Energy Services, L.P.Composition and method for recovering hydrocarbon fluids from a subterranean reservoir
Non-Patent Citations
Reference
1 *HAZOT: "Functionalization of poly (N-methylmethacrylamide) thermosensitive particles by phenylboronic acid", COLLOID POLYMER SCI., vol. 280, 2002, pages 637 - 646
2 *MATSUMOTO: "Glucose-responsive polymer gel bearing phenylborate derivative as a glucose-sensing moiety operating at the physiological pH", BIOMACROMOLECULES, vol. 5, no. 3, 2004, pages 1038 - 1045, XP008159372, DOI: doi:10.1021/bm0345413
3 *NEYRET: "The properties of polyampholyte microgel particles prepared by microemulsion polymerization", POLYMER, vol. 38, no. 25, 1997, pages 6129 - 6134, XP004092650, DOI: doi:10.1016/S0032-3861(97)00173-0
Referenced by
Citing PatentFiling datePublication dateApplicantTitle
WO2010078107A1 *21 Dec 20098 Jul 2010Bausch & Lomb IncorporatedBrush copolymers
WO2011113989A118 Mar 201122 Sep 2011Universidad Del Pas VascoCationic nanogels for biotechnological uses
WO2011159161A220 Jun 201122 Dec 2011Universiteit TwenteBoronated polymers
US841979218 Dec 200916 Apr 2013Bausch & Lomb IncorporatedBrush copolymers
US856923127 Jan 201029 Oct 2013Smartcells, Inc.Soluble non-depot insulin conjugates and uses thereof
US862334527 Jan 20107 Jan 2014SmartcellsTerminally-functionalized conjugates and uses thereof
US884610327 Jan 201030 Sep 2014Smartcells, Inc.Exogenously triggered controlled release materials and uses thereof
US890685027 Jan 20109 Dec 2014Smartcells, Inc.Crystalline insulin-conjugates
US893320722 Jul 201113 Jan 2015Smartcells, Inc.Drug-ligand conjugates, synthesis thereof, and intermediates thereto
US894069027 Jan 201027 Jan 2015National Institutes Of Health (Nih)Synthetic conjugates and uses thereof
US905037027 Jan 20109 Jun 2015Smartcells, Inc.Conjugate based systems for controlled drug delivery
US906801322 Jul 201130 Jun 2015Smart Cells, Inc.Recombinant lectins, binding-site modified lectins and uses thereof
US907401522 Jul 20117 Jul 2015Smartcells, Inc.Recombinantly expressed insulin polypeptides and uses thereof
US91141761 May 201425 Aug 2015Smartcells, Inc.Exogenously triggered controlled release materials and uses thereof
US946324919 May 201411 Oct 2016Smartcells, Inc.Crystalline insulin-conjugates
US95793914 Jun 201528 Feb 2017Smartcells, Inc.Conjugate based systems for controlled drug delivery
Classifications
International ClassificationA61K47/00, G01N21/17, C08J3/075, G01N21/49, A61M5/168, C02F1/26, C02F1/28
Cooperative ClassificationC08J3/075, C02F1/56, A61K47/32, A61K9/06
European ClassificationA61K9/06, A61K47/32, C02F1/56, C08J3/075
Legal Events
DateCodeEventDescription
22 Nov 2006121Ep: the epo has been informed by wipo that ep was designated in this application
2 Oct 2007WWWWipo information: withdrawn in national office
Country of ref document: DE
2 Oct 2007NENPNon-entry into the national phase in:
Ref country code: DE
1 Nov 2007NENPNon-entry into the national phase in:
Ref country code: RU
1 Nov 2007WWWWipo information: withdrawn in national office
Country of ref document: RU
2 Jul 2008122Ep: pct application non-entry in european phase
Ref document number: 06741368
Country of ref document: EP
Kind code of ref document: A1