Microrheological characterization of covalent adaptable hydrogels for applications in oral delivery

Abstract

The feasibility of a covalent adaptable hydrogel (CAH) as an oral delivery platform is explored using μ²rheology, microrheology in a microfluidic device. CAH degradation is initiated by physiologically relevant pHs, including incubation at a single pH and consecutively at different pHs. At a single pH, we determine CAH degradation can be tuned by changing the pH, which can be exploited for controlled release. We calculate the critical relaxation exponent, which defines the gel-sol transition and is independent of the degradation pH. We mimic the changing pH environment through part of the gastrointestinal tract (pH 4.3 to 7.4 or pH 7.4 to 4.3) in our microfluidic device. We determine that dynamic material property evolution is consistent with degradation at a single pH. However, the time scale of degradation is reduced by the history of degradation. These investigations inform the design of this material as a new vehicle for targeted delivery.

Graphical Abstract

μ²rheology, microrheology is a microfluidic device, is used to characterize a covalent adaptable hydrogel scaffold in pH environments that mimic the gastrointestinal tract.

1. Introduction

Covalent adaptable hydrogels (CAHs) are synthetic scaffolds that can rearrange and reform cross-links when pushed out of equilibrium by the incorporation of dynamic covalent chemistries, resulting in responsive structures. At equilibrium, CAHs are stable and behave like conventional covalent cross-linked hydrogels. Bond breakage and reformation is triggered when stimuli pushes the material out of equilibrium^1–7. These environmental and external stimuli include temperature, light, electromagnetic fields, pH and mechanic forces^1,8–16. Due to this, CAHs have a wide range of biological applications including platforms for three-dimensional (3D) cell encapsulation and culture^12,14,17,18, wound healing and tissue regeneration^11,19–23 and drug delivery^{14,20,21,24–27}. As a drug delivery vehicle, active molecules can be released during network degradation. This cargo can also be protected during bond reformation, enabling controlled delivery throughout a biological process. To design the structure of this CAH for these applications, an in-depth understanding of the dynamic Theological properties and microstructure during CAH degradation are crucial.

CAHs are designed using various dynamic covalent chemistries, including Schiff base (imine), Diels-Alder, disulfide and hydra-zone. Each of these chemistries changes the applicability of the CAH and the stimuli they respond to. CAHs developed using imine bonds^9,24,28 and thermally responsive Diels-Alder^13,29,30 reactions have been designed for controlled drug release, cell encapsulation and cell delivery. CAHs designed with dynamic disulfide bonds are sensitive to several stimuli; including light, pH and biological molecules^2,5,31,32, making it ideal for multiresponsive and self-healing applications. Finally, hydrazone bonds have been used as cell culture platforms and tissue mimics^{12,17,26,33,34}. In this work, we characterize a pH-responsive poly(ethylene glycol) (PEG)-hydrazone scaffold developed by McKinnon et al.^12,17,35 focusing on determining whether this scaffold can be tailored for use as a drug delivery vehicle in the gastrointestinal (GI) tract.

The CAH we are characterizing is a hydrogel consisting of 8-arm star PEG-hydrazine and 8-arm star PEG-aldehyde, which create covalent adaptable hydrazone bonds upon mixing^12,17,35,36. Once pushed out of equilibrium, bonds break and reform reversibly. Equilibrium of this scaffold can be shifted by different stimuli including mechanical force and pH^17,35. The pH-responsiveness of this scaffold makes it a material that would undergo dynamic transitions in physiologically relevant environments. Previous studies have characterized the pH-dependent degradation and gelation of this PEG-hydrazone CAH^1,12,17,35. This work indicates scaffold degradation and, subsequently, release of functional molecules from this material can be tuned by changing the environmental pH.

Although the CAH microstructure is recognized as an important design parameter for these scaffolds, previous research has mainly focused on characterizing the change in bulk material properties. In these experiments, the scaffold is characterized using bulk rheology with a stress relaxation experiment^{10,12,13,23,35,37,38}. These experiments apply a large strain to the CAH, effectively breaking all bonds. Then the recovery of the material is measured. If the material returns to the original equilibrium properties it is considered covalent adapt-able^{10,12,13,23,35,37,38}. Beyond stress relaxation experiments, traditional bulk rheological measurements are not effective in characterizing the equilibrium properties of a CAH. Generally covalently cross-linked hydrogels are characterized in the linear viscoelastic regime to obtain equilibrium properties. These measurements would not change the properties of a covalently cross-linked scaffold. But for a CAH the addition of shear to the material would break bonds and change the microstructure. Other characterization includes the adaption of the CAH macroscopic shape when external force is applied^{3,10,12,13,38}. For biological applications both the macroscopic and microscopic evolution of the material is integral in scaffold design. Our work bridges this gap by using passive microrheology to characterize dynamic changes in the material properties and microstructure when the scaffold is degraded by environmental stimuli.

We use multiple particle tracking microrheology (MPT) to characterize our evolving PEG-hydrazone CAH. MPT measures the Brownian motion of fluorescent tracer particles embedded in the material. From the measured Brownian motion, material Theological properties are calculated using the Generalized Stokes-Einstein Relation^41–51. MPT is used to characterize CAH degradation because of several advantages. First, MPT measurements only require small sample volumes, 5μL – 50μL. Second, MPT is sensitive in the low moduli range enabling measurements of the the weak, incipient gel network at critical phase transitions. Third, fast acquisition times of MPT (~30 s) measure the evolving material, with time scales of hours to day, at a quasi-steady state. Finally, MPT uses video microscopy to collect data enabling simultaneous visualization of the spatial material microenvironment. We use MPT to quantitively characterize the dynamic changes in material properties and structure during CAH phase transitions in response to changes in incubation pH. We are specifically interested in how this scaffold changes in pH environments that mimic the GI tract, to determine if this scaffold can be a delivery vehicle for oral drug and nutrient delivery.

In order to mimic pH changes in the GI tract and simultaneously collect MPT measurements of the CAH we use a technique called μ²rheology. μ²rheology is microrheological measurements in a microfluidic device^52–54. To change the pH of the incubation environment around the material, our microfluidic device must enable the exchange of the fluid environment without loss of our CAH. We use a previously designed two-layer microfluidic device to accomplish these environmental pH changes.

In this work, we measure pH-induced degradation of a PEG-hydrazone CAH using μ²rheology, mimicking scaffold degradation in response to pH changes in the GI tract. Two sets of experiment are described. First, scaffold degradation is characterized at a single pH in our microfluidic device. These measurements show that CAH degradation can be tuned by changing pH. Second, the scaffold is degraded consecutively at two different pHs. This is called a consecutive degradation experiment. The change in CAH material properties are consistent with measurements of degradation at a single pH, but the time scale of degradation can be affected by initial scaffold degradation. Characterization of this pH-dependent degradation behavior is crucial in the design of these materials as delivery vehicles. With this information, we can engineer scaffold properties to control the release of molecules from this material enabling targeted delivery of active molecules to an organ and full degradation before exiting the GI tract.

2. Experimental

2.1. Covalent adaptable hydrogel synthesis and sample preparation

The CAH we are characterizing is composed of 8-arm star PEG-hydrazine (M_n 10 000 g mol⁻¹) that covalently cross-links with an 8-arm star PEG-aldehyde (M_n 10 000 g mol⁻¹) creating a covalent adaptable hydrazone network (Figure 1). All scaffolds are made at a 1:1 ratio of hydrazine and aldehyde. The estimated cross-linking efficiency is 41.7%¹⁷. The PEG-hydrazine and PEG-aldehyde are synthesized in the Anseth laboratory at the University of Colorado at Boulder, following previously published protocols, and are the same batches characterized in McKinnon et al.^12,17,35. Briefly, 8-arm star PEG-hydrazine is made by reacting 8-arm PEG end-terminated with amine with activated tri-boc-hydrazinoacetic acid. Tri-boc-hydrazinoacetic acid is activated with hexafluorophosphate azabenzotriazole tetram-ethyl uronium (HATU) and N-methylmorpholine in anhydrous dimethylformamide (DMF). Next, 8-arm PEG-amine in anhydrous DMF is added to the activated tri-boc-hydrazinoacetic acid solution and the reaction proceeds at room temperature overnight. The resulting product is precipitated in ether and the Boc group is removed by dissolving the precipitate in a 50:50 dichloromethane (DCM)-trifluoroacetic acid (TFA) solution for 4 hours. This solution is precipitated into diethyl ether again. The precipitate is dissolved in deionized (DI) water and dialyzing against DI water for 24 hours prior to being lyophilized. Functionality of the macromers is determined by ¹H NMR, which is available in the published literature^12,17,35.

Fig. 1.

Reaction scheme of hydrazone cross-linked PEG molecules.

The 8-arm star PEG-aldehyde is synthesized using a Swern oxidation^56,57. Oxalyl chloride is dissolved in anhydrous DCM and incubated at −78°C in an acetone/dry ice bath³⁵. Diluted dimethyl sulfoxide (DMSO) is added to the oxalyl chloride to form alkoxysulfonium ion intermediate. 8-arm star PEG-hydroxide is dissolved in anhydrous DCM and added to the alkoxysulfonium intermediate solution and reacted for 2 hours. Triethylamine is then added to the resulting solution and allowed to react for 20 mins. Finally, the reaction is warmed to room temperature and the product is precipitated and dialyzed as previously described for the PEG-hydrazine molecule. Functionality of the macromers is determined by ¹H NMR^17,35.

To make our hydrogel samples, 8-arm star PEG-hydrazine, 8-arm star PEG-aldehyde, fluorescent probe particles and buffer solution are mixed. Since gelation is rapid and in order to transfer samples into our microfluidic device prior to gelation, we make hydrogel samples in two steps. The first step mixes 4.4 wt% PEG-aldehyde, 0.04 % solids/volume of 1 μm fluorescently labeled probes particles (2a = 1.0 ± 0.02 μm where a is the particle radius, carboxylated polystyrene probes, Polysciences, Inc.) and buffer solution. Prior to addition to the precursor solution, the probes are washed 3 × using centrifugation and resuspension. They are then sonicated for 15 mins to ensure resuspension and limit aggregation. In these experiments two buffer solutions are used, one is called the ‘acidic buffer’ (pH 4.3) and consists of 74 mM acetic acid and 26 mM sodium acetate. The other is called the ‘physiological buffer’ (pH 7.4) and is made with 200 mM 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid (HEPES) and 1 M sodium hydroxide (NaOH). Buffer solutions are added to the precursor solution to set the pH of gel formation. The buffer pH is adjusted by addition of 1 M hydrochloric acidic (HCl) or 1 M NaOH drop-wise accordingly. The second step in scaffold gelation is to add 4.4 wt% PEG-hydrazine into the first precursor solution and mixing. The mixed precursor solution is immediately injected into our microfluidic device. After mixing the two precursor solutions, the polymers covalently cross-link and form a 3D network. Probe particles are isotropically suspended within this scaffold to enable microrheological measurement. Previous investigations characterized swelling, cell encapsulation, gelation kinetics and viability in this hydrogel scaffold^1,12,17. The rheology of this hydrogel was previously measured and the unswollen equilibrium moduli is reported to be approximately 30 kPa^12,17.

2.2. Multiple particle tracking microrheology

Scaffold degradation is characterized using multiple particle tracking microrheology (MPT), which tracks the movement of probe particles in the material to measure material properties. 1 μm fluorescently labeled carboxylated polystyrene probes (2a =1.0 ± 0.02 μm, Polysciences, Inc.) are embedded in the material prior to gelation. This particle surface chemistry and concentration are chosen to ensure probe particles are not interacting with the polymers in the scaffold and each other, ensuring accurate MPT measurements¹. MPT data are collected on an inverted microscope (Zeiss Observer Zl, Carl Zeiss AG) with a 63 × water immersion objective with a low numerical aperture (N.A. 1.3, lx optovar, Carl Zeiss AG). The Brownian motion of embedded probe particles are recorded using a high-speed camera (Phantom Miro Ml20, 1024 × 1024 pixels, Vision Research Inc.). Videos are collected at 30 frame per second with an exposure time of 1000 μs. These parameters are chosen to minimize static and dynamic particle tracking errors⁴⁵. Particles are tracked by identifying the brightness-weighted centroid of each particle using classic tracking algorithms. The particle centroids in each frame are linked into trajectories using a probability distribution function that accounts for the Brownian motion of a single particle^46–49,58. The ensemble-averaged mean-squared displacement (MSD,〈Δr²(τ)〉) is calculated from these two-dimensional (2D) particle trajectories using the equation 〈Δr²(τ)〉=〈Δx²(τ)〉+〈Δy²(τ)〉, where τ is the lag time, and x and y are position coordinates in the 2D plane. The MSD is directly related to rheological properties, such as the creep compliance (J (t)), using the Generalized Stokes-Einstein Relation

$$\langle \text{Δ}{r}^2\left(\tau \right)\rangle =\frac{k_{b}T}{\pi a}J\left(t\right)$$ (1)

where k_bT is the thermal energy and a is particle radius.

Additionally, the logarithmic slope of the MSD, $\alpha =\frac{d\mathrm{\text{ log }}\mathrm{\langle }\mathrm{\text{Δ}}{\mathrm{r}}^{\mathrm{2}}\mathrm{\left(}\mathrm{\tau }\mathrm{\right)}\mathrm{\rangle }}{\mathrm{d}\mathrm{\text{ log }}\mathrm{\tau }}\mathrm{,}$ identifies the state of the material. When α = 1, probe particles are freely diffusing and the material is a liquid. When α → 0, particles are immobilized in the gel network. When α is between 0 and 1, the material is a viscoelastic sol or gel. The transition between sol and gel is quantitatively defined by comparing α to the critical relaxation exponent, n, which is determined using time-cure superposition.

2.3. Time-cure superposition

Time-cure superposition (TCS) is used to analyze scaffold degradation^{50,51,58–65}. TCS is the superposition of viscoelastic functions at different extents of reaction^{1,50,51,58–63}. Using this analytical method, the critical relaxation exponent, n, and critical degradation time, t_C, are determined. These critical values quantitatively identify the phase transition. During degradation, the phase transition is defined as the transition from the last sample-spanning network cluster in the gel to a sol with no sample-spanning network clusters. When α > n the material is a viscoelastic fluid and when α < n the material is viscoelastic gel. α = n is defined as the critical phase transition, n also identifies the structure of the material at the critical phase transition. n < 0.5 indicates a densely cross-linked network, n > 0.5 indicates an open porous cross-linked network^{1,43,51,58–66}.

To calculate the value of n, MSD curves are shifted on the MSD and lag time axes into pre- and post-gel master curves. MSDs are shifted using their distinct curvature at short lag times, which measures the longest relaxation time of the polymers in the pre-gel and the longest relaxation time of the network in the post-gel^{50,51,58–65} The measured relaxation times change and are distinct at each point during material evolution. Due to this, all MSD curves fit together uniquely into a gel and sol master curve. The slope where the two master curves meet is the gel-sol transition and the critical relaxation exponent, α = n. The value of n is calculated from the shift factors. It should be noted that MSDs are shifted iteratively resulting in shift factors that can be slightly different in the vertical and horizontal direction each time master curves are made. Although the absolute values can be slightly different these changes in the shift factors are related, and the same value of n is always calculated.

The lag time is shifted by shift factor a, which relates to the longest relaxation time, τ_L, and the distance away from critical extent of degradation, $\mathcal{E}=\frac{\left|t-t_c\right|}{t_c},$ by scaling exponent y using the relation

$$a~{\tau }_L^{-1}~{\left(\frac{\left|t-t_c\right|}{t_c}\right)}^y.$$ (2)

MSDs are shifted by shift factor b, which relates to the steady state creep compliance, J_e, and the distance away from critical extent of degradation by scaling exponent z using the relation

$$b~J_e^{-1}~{\left(\frac{\left|t-t_c\right|}{t_c}\right)}^z.$$ (3)

The ratio of the scalings exponent is the critical relaxation exponent^{50,51,59–60}

$$n=\frac{z}{y}.$$ (4)

All values of n are reported as n_avg, which is the average value of at least three experiments. The error reported is the propagation of the errors from these experiments.

2.4. Microfluidic device fabrication and μ²rheology

To mimic pH changes in the body, specifically in the digestive tract, we measure the change in material properties of our CAH in a two-layered microfluidic device that changes the fluid environment around a sample^52,53,55. This device creates equal pressure around the sample during fluid exchange, which traps the material in place enabling minimal sample loss^52,55. Therefore, we can exchange the fluid environment around our CAH several times while characterizing scaffold degradation. After fluid exchange, MPT measurements are taken with no flow in the sample chamber.

In Figure 2, the hydrazone CAH is injected into the sample chamber (the first layer of the device) which is below the solvent basin (second layer of the device). The solvent basin is connected to the sample chamber by six channels. These channels are evenly spaced, every 60°, around the edge of the sample chamber to create equal pressure. After exchange of the fluid microenvironment, MPT measurements are collected of the evolving CAH. These MPT measurements in the microfluidic device are μ²rheology measurements.

Fig. 2.

Microfluidic device used for μ²rheology measurements. The top image is a picture of the top of the device and a diagram is provided with a side view of the first layer (middle image) and the second layer (bottom image). The CAH is injected into the sample chamber (first layer of the device), which is below the solvent basin (second layer of the device). The solvent basin connects to the sample chamber through six channels that are evenly distributed around the edge of the chamber. When fluid is exchanged using suction through the suction chamber, the fluid enters the sample chamber through these six channels, creating equal pressure in the sample chamber that traps the CAH in place. This results in complete fluid exchange with minimal sample loss.

The device fabrication and theory of operation are described in detail in recent publications^52,53,55. Here, we briefly describe key aspects of the fabrication and operation of this device. To fabricate the device, first the channel design is printed on a transparency (8.5 × 11 in, Apollo) called the design negative. Next, a stamp is made of the channel design cured onto a glass slide in ultraviolet (UV) resin. To create the stamp a stack of three layers is made prior to polymerization. The base of the stack is a blank transparency. On the transparency four glass spacers (cut from a 25 × 75 × 1 mm glass slide, Fisher Scientific) are placed to set the height of the stamp and UV curable thiolene resin (NOA 81, Norland Products, Inc.) is poured between the spacers. The second layer is a glass slide (75 × 50 × 1 mm, Fisher Scientific), which contacts the resin and sits on the spacers. The third layer is the design negative that it is placed on top of the glass slide and exposed to UV light (Spectroline SB-100P, 365 nm). After exposure for 45 s, the design is polymerized onto the glass slide with a height of 1 mm. Next, the excess resin on the stamp is removed using acetone (VWR Analytical), ethanol (%, Fisher Scientific) and water. The stamp is post-cured for 30 mins under UV light to ensure complete polymerization.

To create the microfluidic device, the stamp is placed into a petri dish (150 × 15 mm, Falcon) and polydimethylsiloxane (PDMS, Sylgard 184, DOW Corning) is poured over the stamp and cured at 55°C overnight. The PDMS is made by fully mixing a 1:10 ratio of curing agent to silicone elastomer base. After curing, the patterned PDMS is cut out and sealed to a 75 × 50 × 0.15 mm glass slide (Fisher Scientific) using plasma treatment. Due to the porous nature of PDMS, the inner channel surfaces are coated with a glass layer using sol-gel chemistry. A tetraethoxysilane (TEOS) and methyltriethoxysilane (MTES) precursor solution is heated to 100°C for 10 s to create a thin glass layer inside the channels of our device^52,67,68. This glass layer prevents sample uptake into the PDMS during long experiments. Finally, the solvent basin is made using unpatterned PDMS and is sealed to the the first layer above the sample chamber using plasma treatment, making a two-layer device^52,53,55.

To measure degradation at a single pH, the buffer solution is first loaded into the device and then the hydrogel sample is injected into the sample chamber.μ²rheology measurements are then collected of the CAH evolution in this pH environment. To measure CAH degradation after a single exchange of buffer solution, the material is first degraded at a single pH as described above. The buffer is then exchanged by continually adding new pH buffer solution into the solvent basin which enters the sample chamber and removing buffer from the sample chamber by pulling gentle suction through the suction chamber using a syringe pump at a constant flow rate of 1 mL mins⁻¹. This enables the second incubation buffer to be pulled into the sample chamber to begin CAH degradation in the new pH environment. In order to ensure the buffer is completely exchanged, the total volume exchanged is at least 5× the original buffer volume. The same procedure is followed for both experiments outlined above. The buffer in the sample chamber is exchanged 3 × to ensure that scaffold degradation is measured in the new pH environment.

2.5. CAH degradation at a single pH

CAH degradation is induced by incubation in either acidic (pH 4.3), where the gel is formed at physiological pH (pH 7.4), or incubation in physiological (pH 7.4), where the gel is made at acidic pH (pH 4.3), pH buffer solutions. The change in pH pushes the hydrazone bonds out of equilibrium, where both bond degradation and formation occur. Here, we mimic the environments of the stomach (pH 4.3) and the intestines (pH 7.4). pH 4.3 is chosen to represent the pH of the stomach. This is an average value chosen considering the pH range in the upper and lower stomach and the change in pH during digestion. pH 7.4 is a physiologically relevant pH and is found in the intestines^69,70. Throughout the experiment there is no flow in the sample chamber but the solvent basin is continually filled with buffer solution, preventing any effects from solvent evaporation. Data of acidic degradation are collected every 5 – 6 mins over the course of the entire degradation reaction, which takes approximately 10 hrs. Complete degradation at physiological pH takes approximately 10 days, and data are collected every 10 mins for 3 – 5 hrs each day.

Prior to measuring degradation in the microfluidic device, we characterize scaffold degradation at a single pH in glass-bottomed petri dishes (MatTek Corporation, D = 35 mm). These results are compared with degradation at a single pH in our microfluidic device, shown in supplemental Figure S1^† These experiments were done to ensure that the microfluidic device does not change CAH degradation. The fabrication of the sample chambers in the glass-bottomed petri dishes is described in recent publications^50,71. Briefly, the sample chamber is a tube of cured PDMS (Dow Corning, O.D. 10 mm and I.D. 6 mm) attached to the functionalized glass-bottom of the petri dish using uncured PDMS. The glass-bottom is functionalized with (3-mercaptopropyl)triethoxysilane (Sigma-Aldrich Co. LLC) to enable the hydrogel sample to chemically attach to the glass, immobilizing the CAH during degradation. The hydrogel precursor is mixed in two steps as described above and quickly injected into the PDMS sample chamber. The material instantly gels and is then incubated in a buffer solution. MPT measurements are collected as described above.

2.6. Consecutive degradation at different pHs

For consecutive degradation experiments, we measure degradation at acidic (pH 4.3) and physiological (pH 7.4) pHs. In the first experiments, scaffolds are made at pH 7.4 and first incubated at pH 4.3. At pH 4.3 the material undergoes cycles of degradation followed by spontaneous re-gelation, where α → 0. After re-gelation, the buffer is exchanged to pH 7.4 until complete scaffold degradation is measured. This degradation scheme mimics the pH environment in the GI tract starting in the mouth through the stomach to the large intestines. In the second experiment, the scaffold is made at pH 4.3, the pH of the stomach, and first degraded at pH 7.4 until α ≈ 0.7. The buffer is then exchanged to pH 4.3 until complete CAH degradation. This pH variation mimics degradation in the duodenum and the large intestines.

3. Results and discussion

We use μ²rheology to characterize CAH degradation initiated by changes in environmental pH mimicking the GI tract. The overall goal is to determine the viability of this CAH as a delivery vehicle. First, CAH degradation is measured at a single pH. These results are compared to previous and control MPT experiments to ensure that the microfluidic device does not change material degradation¹. Next, we characterize scaffold degradation when it is degraded at two different pHs consecutively. These measurements determine the time of degradation and structure of this CAH as the pH of the digestive tract is mimicked. With this information, the design of the structure of this material can be tuned to make it an effective vehicle for oral delivery of active molecules, such as drugs and nutrients.

3.1. CAH degradation at a single pH

CAH degradation is characterized at a single pH in our microfluidic device. Figure 3a and b are the logarithmic slopes of the mean-squared displacement, α, as a function of CAH degradation time in our microfluidic device at pH 7.4 and 4.3, respectively. The error bars on α for both graphs are the standard deviations of fitting an MSD curve to a straight line.

Fig. 3.

CAH degradation at a single pH. The logarithmic slope of the mean-squared displacement during degradation at (a) physiological (pH 7.4) and (b) acidic (pH 4.3) pH. The early data acquisition times at pH 7.4 are shown as an inset in (a). The dashed line is the critical relaxation exponent, n, where the gel-sol transition occurs, n is determined using time-cure superposition.

In Figure 3a at t = 0 mins, α = 0 indicating that immediately after gelation the scaffold is a gel. As time increases, α increases indicating the structure of the scaffold network is degrading and probe particle movement is increasing. The slope continues to increase rapidly during initial degradation and increases past the critical relaxation exponent, n_avg = 0.79 ±0.06 (dashed line in Figure 3a) at t = 5025 mins, which is calculated using TCS. This indicates that the material has transitioned from a gel to a sol. The determination of n is discussed in detail below. After the gel-sol transition, the properties of the scaffold oscillate around the critical transition point until α > 0.8, indicating that the material has transitioned to a sol, around 7100 mins. During these oscillations the material is forming and breaking a sample-spanning gel network prior to complete degradation. Complete degradation takes approximately 10 days, which is not shown in Figure 3a. These measurements follow the same degradation trends as previous investigations and the value of n measured in the microfluidic device, n_avg = 0.79 ± 0.06, is within error of a previously published value, n_avg = 0.86 ± 0.04¹. Therefore, there is no change in CAH degradation in the microfluidic device at physiological pH¹.

Scaffold degradation at acidic pH is shown in Figure 3b. At t = 0 mins, μ²rheology measures α = 0 indicating that the material is a gel. As time increases, the slope increases indicative of scaffold degradation. The CAH quickly degrades to α ≈ 0.7 after approximately 100 mins. At this point, the slope quickly decreases to α ≈ 0.1, indicating that the material has spontaneously re-gelled. It should be noted, that no additional stimuli is added to the system to cause the material to re-gel. After the CAH re-gels, the scaffold starts to degrade again until spontaneous re-gelation is measured. This is called a ‘degradation-gelation cycle’. In Figure 3b, the dashed line is the critical relaxation exponent, n. We measure values of α ≥ n during each cycle. This indicates that the material is reaching or undergoing a phase transition during each degradation-gelation cycle. At least two cycles are measured before complete degradation, t > 10 hours, and the time period of each degradation-gelation cycle is measured to be approximately 250 mins. Complete material degradation occurs due to a decrease in polymer concentration from diffusion of polymers away from the scaffold when the material is a sol.

Degradation-gelation cycles at pH 4.3 are measured throughout the gel. This is evidenced by our MPT measurements. All measurements are a single measurement of many particles, therefore, we are measuring the change in properties in the entire field of view. Ensemble-averaged van Hove correlations have also been calculated for particles in this material and have confirmed all particles are experiencing the same microenvironment. In addition, measurements in macroscopically distinct environments measure the same phenomenon.

From these measurements we conclude that this degradation reaction is pH dependent. In our scaffold the hydrazone bond undergoes a proton exchange reaction, which is the source of the measured pH dependent degradation. At acidic pH, due to the excess of protons in solution the equilibrium is strongly shifted towards degradation. After almost complete degradation, the equilibrium shifts again to bond formation. This is the basis of the degradation-gelation cycles. At pH 7.4, equilibrium is shifted due to a change in pH from the pH of gel formation. Since there is not an excess of protons in solution at pH 7.4 the reaction equilibrium is shifted towards bond hydrolysis but there is also simultaneous bond formation, which is evident in oscillations around the gel-sol transition.

A previous study described hydrazone degradation with first-and second-order reaction kinetics at different pHs using MPT measurements. The reaction constants from MPT agree with reaction constants measured using small molecule kinetics when the different systems measured are accounted for^1,35. At pH 4.3, the reaction rate constant of hydrazone bond hydrolysis from McKinnon et al. is k₋₁ = 0.76 mins⁻¹. It is in the same order of magnitude with the reaction rate constant measured at pH 4.3 from MPT data k₋₁ = 0.16 ±0.02 mins⁻¹. At pH 7.4, McKinnon measures k₋₁ = 0.03 mins⁻¹ and from MPT k₋₁ = 0.003 ± 1.8 × 10⁻⁵ mins⁻¹. This is an order of magnitude difference and can be explained by the different systems that are measured. McKinnon’s experiments only use the functional groups of the scaffold. MPT characterizes hydrazone cross-linked hydrogels consisting of 8-arm PEG-aldehyde and 8-arm PEG hydrazine.

Additionally, polymer diffusion out of our scaffold should also be taken into account, especially when the material transitions from a gel to a sol. At the phase transition polymers are no longer connected to the network and can diffuse away from the gel, those polymers will be washed out of the system during buffer exchange. At this point they no longer contribute to the elasticity of the network and also cannot participate in network bond formation.

3.2. Time-cure superposition

The critical values at the phase transitions, such as the critical relaxation exponent, n, and critical degradation time, t_c, are determined using time-cure superposition^{1,50,51,58–63}. An example of TCS for physiological degradation is provided in Figure 4. MSDs and shifted master curves are in Figures 4a and b, respectively, the slope at the intersection of two master curves is the value of n.

Fig. 4.

Time-cure superposition of CAH degradation at physiological pH.(a) Measured mean-squared displacement versus lag time, color represents the degradation time, (b) Measured MSDs are shifted into sol and gel master curves, (c) Shift factors a and b go to zero at the critical degradation time, t_c = 5000 mins. (d) The critical relaxation exponent, n, is calculated from the shift factors, and the average value of n for this experiments is n_avg = 0.76 ±0.07. Three physiological degradation replicates result in a critical relaxation exponent of n_avg = 0.79 ±0.06, which is within error of values calculated for degradation at pH 4.3.

Shift factors go to zero at the critical degradation time, t_c, defined as the time when the last sample-spanning network cluster breaks, Figure 4c. This is because the elastic moduli in the gel goes to zero and viscosity in the sol goes to infinity at the critical phase transition. In Figure 4d, the logarithm of the shift factors, In a and In b, versus the logarithm of the distance from the critical degradation time, In ε, is fit to a line (represented using a dashed and dotted line). This is done because the shift factors are related to the longest relaxation time and steady state creep compliance, which exhibit power law behavior before and after the gel point as described in Equations 2 and 3. The critical scaling exponents y and z are calculated from this fit and used to calculate the critical relaxation exponent, $n=\frac{z}{y}.$ TCS for acidic degradation is provided in the Supplementary Information in Figures S2–S3^†.

The critical relaxation exponent is a material property which is a quantitative measure of the scaffold structure similar to a complex modulus, G*. High values of n (i.e. 0.5 <n< 1) indicate an open porous cross-linked network which dissipates more energy than it stores. A low value of n (i.e. 0.1 < n < 0.5) indicates a densely cross-linked network that store more energy than it dissipates^{1,43,51,58–66}. At physiological pH, n_avg = 0.79 ±0.06, and acidic pH, n_avg = 0.73 ±0.08, these materials have a loosely cross-linked network at the critical transition.

In the microfluidic device, degradation of the CAH follows the same trends in material properties and critical values measured in control experiment. The control experiment in acidic buffer is shown in Figure S1^† The control experiment in physiological buffer has been published previously¹. Experiments in acidic buffer do not agree with previously published results because the buffer solution was changed because the previous buffer had side reactions with the functional groups which changed the initial gel structure. At pH 4.3 in the microfluidic device, n_avg = 0.73 ±0.08, and in control experiments the value is n_avg = 0.78 ±0.08 (Figure S1^†). There is also agreement between the value of n_avg in physiological pH buffer. Additionally, all values of n_avg are within error of each other. As mentioned previously, n is a material property and should only change when the structure of the gel changes. A structural change would occur by changing the number of functional groups or the size of the polymers^{1,50,51,58–63}.

3.3. Consecutive CAH degradation at physiological and acidic conditions

Consecutive degradation from physiological pH to acidic pH mimics the pH from the duodenum (pH 7.4) through the small intestines (pH 4.3) to the large intestines (pH 4.3). The small intestines is where most chemical digestion takes place and is a target site for drug and nutrient absorption and transport^72,73. We characterize scaffold degradation and material properties and determine if the evolving scaffold is dependent on the degradation history.

Our first experiments degrade the scaffold at physiological pH until the value of α ≈ 0.7, where the material passes the gel-sol transition. When the scaffold degrades to this point, the buffer is exchanged to acidic buffer in the sample chamber of the microfluidic device. The scaffold is then degraded in acidic conditions until complete degradation. MPT measurements of this degradation scheme are shown in Figure 5a–b. Figure 5a is the measured mean-squared displacement of a consecutive degradation reactions. The color of the MSDs represents the degradation time.

Fig. 5.

Consecutive CAH degradation at physiological and then acidic conditions, (a) Measured mean-squared displacement of all measurements versus lag time, color represents degradation time, (b) Logarithmic slope of the MSD throughout the entire degradation experiment. The vertical dotted line represents the time when buffer is exchanged from physiological to acidic pH. The horizontal dashed line is the n value, which indicates the phase transitions. This value of n is determined for this consecutive degradation experiment using time-cure superposition.

In Figure 5a and b, we initially measure steady degradation at physiological pH, where the scaffold quickly passes the gel-sol transition prior to buffer exchange. When the CAH degrades to a slope of approximately 0.7, the buffer solution is exchanged to acidic pH, indicated by the vertical dashed line in Figure 5b. We continue to measure changes in CAH material properties at acidic conditions. At pH 4.3, we again measure degradation-gelation cycles, similar to the cycles measured in control experiments. Even after the scaffold has undergone the gel-sol transition at pH 7.4, the CAH still undergoes degradation-gelation cycles when the incubation environment is changed to pH 4.3.

Consecutive degradation at two different pHs, pH 7.4 and then pH 4.3, is compared to single pH experiments in the microfluidic device. As seen in Figure 6a and b, the change in material properties is consistent with degradation at a single pH. Degradation within the microfluidic device at pH 7.4 is not changed, Figure 6a. More interestingly, CAH degradation at pH 4.3 is unchanged by initial degradation at pH 7.4. In Figure 6b, the time of a degradation-gelation cycle, defined as the time between complete gelation events, is noted on the graph. For the single and consecutive degradation experiments the time is Δt = 316 mins and Δt = 364 mins, respectively These times are similar indicating that the initial degradation at pH 7.4 has no effect on the change in material properties. This can be explained by the kinetics of degradation at pH 7.4. At physiological pH, scaffold degradation is a first-order reaction, the reaction equilibrium is shifted towards hydrolysis but does have simultaneous bond formation¹. Because of this there is minimal loss of polymer from diffusion out of the scaffold during the initial gel-sol transition at pH 7.4. Therefore, the CAH scaffold has a similar cross-link density when it re-gels after the buffer is exchanged to acidic pH.

Fig. 6.

Comparison of single and consecutive CAH degradation. The change in the logarithmic slope of the MSD, $\alpha =\frac{d\mathrm{\text{ log }}\mathrm{\langle }\mathrm{\text{Δ}}{\mathrm{r}}^{\mathrm{2}}\mathrm{\left(}\mathrm{\tau }\mathrm{\right)}\mathrm{\rangle }}{\mathrm{d}\mathrm{\text{ log }}\mathrm{\tau }}\mathrm{,}$ is plotted versus normalized degradation time at (a) physiologicaf and (b) acidic pH. The trend in material properties and time scale of degradation remain consistent between all measurements.

Figure 7a and b is analysis of this experiment using time-cure superposition. With this analysis, we confirm that there is no change in the structure of the material or the critical transition point when the scaffold is degraded consecutively at two different pHs. We use data from degradation at both pHs to create one sol and one gel master curve, Figure 7a. This is done because the material properties and longest relaxation times measured should be consistent regardless of the incubation pH. Figure 7b measures a gel-sol transition at pH 7.4 and two sol-gel and one gel-sol transition at pH 4.3. n is calculated for the pre- and post-gel for each degradation measured. The average value of n is n_avg = 0.64 ± 0.09 for this experiment. This value of n is within error of the n_avg calculated for CAH degradation at a single pH, n_avg = 0.76 ±0.05. The values of n have also been analyzed using a t-test with p < 0.05. From this analysis, the difference between the control and consecutive degradation n values is not statistically significant. This result further confirms that pH doesn’t change the structure or properties of the material at the critical transition.

Fig. 7.

Time-cure superposition of consecutive degradation at physiological and acidic pH. (a) Measured MSDs are shifted into sol and gel master curves, (b) The shift factors a and b go to zero at each critical degradation time, where t_c1 = 146 mins, t_c2 = 214 mins, t_c,3 = 454 mins and t_c4 = 477 mins. This experiment results in a critical relaxation exponent of n_avg = 0.64 ±0.09, which is within error of values calculated for single pH experiments.

This experiment mimics the change in pH the CAH would experience when it transitions from the duodenum to the small intestines. In the duodenum, the scaffold would degrade through the gel-sol transition. The passage to the small intestines is mimicked by the exchange of the buffer environment to pH 4.3. The material would start as a sol in the small intestines, which could enable release of cargo and could increase delivery efficiency. Once cargo is released in the small intestines, the spontaneous gelation event would retain remaining cargo within the scaffold for release later as it moves through the intestines. Although these experiments do not mimic the temporal changes in the GI tract, they provide information about the changes in the state and material properties of the CAH as it moves through these different environments. This information can be used to tailor the material for sustained release of functional molecules within the therapeutic window while minimizing side effects, such as blockage⁷².

3.4. Consecutive CAH degradation at acidic and physiological conditions

We now characterize consecutive CAH degradation at acidic and then physiological pH. In an acidic buffer environment, the scaffold is degraded until it undergoes at least two complete degradation-gelation cycles. When the CAH re-gels the second time, the buffer is exchanged to physiological pH. This degradation mimics part of the pH environment in the GI tract, specifically degradation from the stomach (pH 4.3) through the duodenum and the intestines (pH 7.4). Again it should be noted that the pH environment in the intestines ranges from pH 4 – 7⁷².

Figure 8a is the logarithmic slope of the MSD for this degradation scheme. The vertical dashed line represents the time when the buffer is exchanged from acidic to physiological pH. At pH 4.3, we measure cycles of degradation followed by almost complete spontaneous gelation. Once the buffer is exchanged to pH 7.4, we continue to measure CAH degradation without re-gelation events. At pH 7.4, the scaffold initially degrades to the gel-sol transition and then oscillates around the critical transition point before completely degrading. The results from consecutive degradation are compared to control experiments of CAH degradation at a single pH. We determine that the change in material properties is in good agreement with measurements of degradation at a single pH. This data is provided in the Supplementary Information in Figures S6b–e^†.

Fig. 8.

Consecutive degradation at acidic and physiological pH. (a) MPT measures the change in $\alpha =\frac{d\mathrm{\text{ log }}\mathrm{\langle }\mathrm{\text{Δ}}{\mathrm{r}}^{\mathrm{2}}\mathrm{\left(}\mathrm{\tau }\mathrm{\right)}\mathrm{\rangle }}{\mathrm{d}\mathrm{\text{ log }}\mathrm{\tau }}$ versus time. The vertical dotted line indicates the change in buffer from pH 4.3 (left) and pH 7.4 (right). The early data acquisition times at pH 4.3 are shown as an inset in (a),(b) Comparison of consecutive and single CAH degradation at pH 7.4. At physiological conditions, the degradation time is significantly decreased after the scaffold is initially degraded in acidic buffer. The time of complete degradation is reduced from 10 to 3 days. The comparison at acidic pH is provided in the Supporting Information, Figure S6^†

In this experiment, we do measure a change in the time scale of degradation after the buffer environment is changed to pH 7.4. No change in time of the degradation-gelation cycles are measured at pH 4.3, Figure S6^†. Figure 8b is a comparison of degradation at physiological pH for our single and consecutive degradation experiments. We measure a significant reduction in CAH degradation time at physiological pH when the scaffold is first degraded in acidic buffer. Complete degradation occurs in three days. However, for scaffolds only degraded in physiological pH, we measure complete degradation after 1.5 weeks. The time scale of physiological degradation is greatly reduced because there is loss of polymer due to diffusion during the initial degradation at pH 4.3. This results in a network with a lower cross-link density when degradation at physiological pH begins. Since cross-link density is reduced, the scaffold degrades (α →1) faster at physiological pH.

Time-cure superposition is also used to analyze this consecutive degradation. Figure 9a is the degradation profile for the first day of the experiment. This graph includes measurements of degradation at both acidic and physiological pH. Two master curves are created using all measurements, shown in Figure 9b. In Figure 9c, seven phase transitions have been characterized in this experiment. From the shift factors, the critical relaxation exponent is calculated, n_avg = 0.69 ±0.11. Similarly, this value of n is also within error of our single pH degradation experiments and agrees with the value calculated for our previous consecutive degradation experiment. A t-test for all n values, p < 0.05, shows no statistically significant difference for degradation at a single pH and consecutive degradation at two pHs.

Fig. 9.

Consecutive CAH degradation in physiological and acidic conditions. $\alpha =\frac{d\mathrm{\text{ log }}\mathrm{\langle }\mathrm{\text{Δ}}{\mathrm{r}}^{\mathrm{2}}\mathrm{\left(}\mathrm{\tau }\mathrm{\right)}\mathrm{\rangle }}{\mathrm{d}\mathrm{\text{ log }}\mathrm{\tau }}$ versus time for the first 24 hrs of the experiment. The vertical dotted line indicates when buffer is exchanged. The critical relaxation exponent is determined using TCS, which is labeled by the horizontal dashed line, and indicates the CAH undergoes multiple gel-sol and sol-gel transitions, (b) Measured MSD in the first day of degradation is shifted into sol and gel master curves, (c) Shift factors go to zero at each critical degradation time, which are t_c,1 = 118 mins, t_c,2 = 120 mins, t_c,3 = 239 mins, t_c,4 = 242 mins, t_c,5 = 316 mins, t_c,6 = 348 mins and t_c,7 = 394 mins. The critical relaxation exponent is n_avg = 0.69 ±0.11 agrees with all previous measurements of this scaffold.

This degradation profile gives more information about the potential of this CAH as an oral delivery vehicle. For targeted therapeutic delivery, we want to minimize cargo release in the stomach and maximize release in the small intestines. In the stomach (pH 4.3), the scaffold undergoes phase transitions from the gel to the sol, which is followed by rapid re-gelation. These degradation-gelation cycles could prevent the scaffold from prematurely degrading and minimize release in the stomach. After the CAH passes to the duodenum and intestines, which are at physiological pH, the scaffold will quickly degrade to a sol followed by oscillations around the critical transition, which could enable controlled, sustained release of active molecules in the intestines within the therapeutic window^72–74.

4. Conclusions

In this study, CAH degradation is characterized as a function of environmental pH, including at a single pH and consecutive degradation at different pHs. Using μ²rheology, we mimic pH changes in the gastrointestinal tract and characterize scaffold degradation in these physiologically relevant environments. Initial experiments characterize CAH degradation at a single pH within our microfluidic device. We determine that the scaffold has an open porous structure at the critical transition, which is a material property and is not dependent on the pH that initiates degradation. Experiments are then designed to exploit the functionality of the microfluidic device and mimic pH changes in the GI tract. These experiments characterize the CAH during consecutive degradation at different pHs. The change in material properties during consecutive scaffold degradation at two different pHs is consistent with measurements of degradation at a single pH. Although the evolution of material properties is unaffected by the pH of the incubation environment, we do measure a time scale change in one of our consecutive degradation schemes. For a scaffold initially degraded in physiological pH then changed to acidic pH, the time scale of material degradation is unchanged. For a scaffold initially degraded in acidic pH and then changed to physiological conditions, the time scale of physiological degradation is greatly reduced. From μ²rheology, we determine that this CAH has distinct changes in the evolution of material properties as a result of the pH of the incubation environment, which can be exploited to design these materials as delivery vehicles for oral drug and nutrient delivery.

In this work, we explore the capability of this CAH as an oral delivery platform and expand the use of μ²rheology to mimic the complex pH-dependent digestive system. These results suggest that this CAH can be tailored to be an effective delivery vehicle. Future work will expand the knowledge base of the evolution of this material by using μ²rheology to mimic temporal pH changes through the entire GI tract, necessitating several buffer exchanges. This technique can also be expanded to include enzymatic degradation to further mimic native digestion. This work suggests that material properties can be engineered, including changing the polymer concentration and functionality, to create a scaffold that will degrade and simultaneously release active molecules for targeted delivery.