Diaphragmatic recovery in rats with cervical spinal cord injury induced by a theophylline nanoconjugate: Challenges for clinical use

Abstract

Context: Following a spinal cord hemisection at the second cervical segment the ipsilateral hemidiaphragm is paralyzed due to the disruption of the rostral ventral respiratory group (rVRG) axons descending to the ipsilateral phrenic motoneurons (PN). Systemically administered theophylline activates a functionally latent crossed phrenic pathway (CPP) which decussates caudal to the hemisection and activates phrenic motoneurons ipsilateral to the hemisection. The result is return of function to the paralyzed hemidiaphragm. Unfortunately, in humans, systemically administered theophylline at a therapeutic dose produces many unwanted side effects.

Design and setting: A tripartite nanoconjugate was synthesized in which theophylline was coupled to a neuronal tracer, wheat germ agglutinin conjugated to horseradish peroxidase (WGA-HRP), using gold nanoparticles as the coupler. Following intradiaphragmatic injection of the nanoconjugate, WGA-HRP selectively targets the theophylline-bound nanoconjugate to phrenic motoneurons initially, followed by neurons in the rVRG by retrograde transsynaptic transport.

Participants: (N/A)

Interventions: (N/A)

Outcome Measures: Immunostaining, Electromyography (EMG).

Results: Delivery of the theophylline-coupled nanoconjugate to the nuclei involved in respiration induces a return of respiratory activity as detected by EMG of the diaphragm and a modest return of phrenic nerve activity.

Conclusion: In addition to the modest return of phrenic nerve activity, there were many difficulties using the theophylline nanoconjugate because of its chemical instability, which suggests that the theophylline nanoconjugate should not be developed for clinical use as explained herein.

Introduction

Theophylline (1, 3-dimethylxanthine), a methylxanthine, acting as a bronchodilator, has been clinically prescribed to treat symptoms of asthma, bronchitis, and emphysema.^1,2 Specifically, theophylline has been shown to enhance respiratory drive by acting as an adenosine receptor antagonist^3–5 that blocks both A₁ and A₂ adenosine receptors.⁶ In C2Hx (C2 hemisected) injured rats, systemic (oral or intravenous) administration of theophylline stimulates the CPP, which activates phrenic motoneurons ipsilateral to the hemisection (Fig. 1), by increasing respiratory output; this results in recovery of function of both the phrenic nerve and diaphragm.^6,7 Theophylline-induced return of diaphragm function can be transient or permanent based on the number and frequency of doses (Nantwi et al.,⁶ 1x intravenous; Nantwi et al.,⁸ multiple oral doses).

Figure 1.

Respiratory pathway involved in the crossed phrenic pathway (CPP). The spinal cord hemisection was induced at C2. The arrows showed the pathway activated phrenic motoneurons ipsilateral to the hemisection, and resulted in recovery function of both the phrenic neuron and diaphragm.

Clinical use of theophylline in man has been reported as early as 1922.⁹ However, there is limited documentation of clinical studies that investigate the use of systemically administered theophylline in humans following SCI to increase respiratory drive.^10–12 In humans, the therapeutic dose of methylxanthines, including theophylline, can cause intolerable side effects including nausea, vomiting, nervousness, increased or irregular heartbeat, and insomnia.^1,12,13 The side effects are caused by a high concentration of the drug in the plasma leading to global phosphodiesterase inhibition and adenosine A₁-receptor antagonism resulting in a hyperactive state of many non-respiratory neurons.¹³ Further efforts to investigate the clinical outcome of systemic theophylline administration have been abandoned due to intolerable side effects and lack of quality data. An alternative method to administer theophylline is also needed. The ability to target theophylline to select respiratory nuclei has the potential to greatly reduce the systemic therapeutic dose and to reduce or eliminate unwanted side effects.

Based on the demonstration of transsynaptic transport of WGA-HRP in the phrenic motor system following a C2Hx,¹⁴ WGA-HRP was utilized to target drug delivery to the motoneurons of the phrenic nucleus (PN) and the pre-motor rostral ventral respiratory group (rVRG) neurons involved in diaphragm function. It was demonstrated that the WGA-HRP was retrogradely transported to the PN and then to the rVRG over the CPP.¹⁴ To bind theophylline (or potentially any drug or molecule of choice) to WGA-HRP, a coupler (or carrier) was needed to facilitate the chemical conjugation. In vivo application of gold nanoparticles (AuNPs) has been established in the literature and are known for their biocompatibility, low toxicity, and easy attachment of various structures via chemical bonds to the AuNPs.^15–24 Following injection into the left hemidiaphragm (LHD, paralyzed by the C2 hemisection), the synthesized tripartite nanoconjugate, comprised of WGA-HRP chemically conjugated to AuNPs which in turn is chemically conjugated to theophylline, undergoes WGA-HRP-receptor-mediated endocytosis. Once endocytosed the nanoconjugate is transported in a retrograde manner to the PN in the spinal cord followed by retrograde transsynaptic transport to the rVRGs in the medulla.²⁵ The ester bond linking the AuNP to theophylline is biodegradable, designed to release theophylline following endocytosis.²⁵ The expected result is the recovery of the hemidiaphragm previously paralyzed by the C2Hx.

Prior to this study a similar nanoconjugate, coupled to 1,3-Dipropyl-8-cyclopentylxanthine (DPCPX) in place of theophylline, was investigated in the same C2 spinal hemisection model. The DPCPX-coupled nanoconjugate produced a significant amount of phrenic nerve recovery in rats under standardized conditions; 56.8 ± 4.3% (1 week), 72.4 ± 7.3% (2 weeks), 46.5 ± 11.8% (4 weeks), compared to the controls at 8.2 ± 2.3%; left phrenic nerve activity expressed as a percentage of right phrenic nerve activity.²⁶ Currently DPCPX is an FDA approved compound for cystic fibrosis in humans²⁷ whereas theophylline has been approved after SCI^10–12 and therefore theophylline was our focus in this study. That is, to use theophylline as the recovery-inducing drug in the tripartite nanoconjugate instead of DPCPX. As we will show, there are many problems with the in vivo use of the theophylline nanoconjugates. Furthermore, with the modest phrenic nerve recovery induced by theophylline nanoconjugates, (compared to the recovery induced by DPCPX nanoconjugates), we suggest that theophylline nanoconjugates should not be pursued for clinical use.

Results

Immunostaining

Three rats underwent intradiaphragmatic injections of the nanoconjugate following a C2Hx to visualize the locations of the nanoconjugate. An immunohistochemical technique was utilized to visualize the WGA component of the nanoconjugate. The WGA-HRP-AuNP bond is a relatively long-lasting amide bond whereas the AuNP-proTHP bond is a transient ester bond. Therefore, we can assume that the location of WGA should also identify where the AuNPs are but not necessarily the proTHP. All 3 rats produced identical results verifying what was reported in.²⁵ WGA positive label was identified in the ipsilateral PN in the cervical spinal cord and bilaterally in the rVRGs in the medulla (Fig. 2). The bilateral rVRGs labeling was notably less intense compared to the label in the PN. These results suggest that at minimum, the WGA-HRP-AuNP portion of the nanoconjugate is capable of being transported to the PN and subsequently transsynaptically to the rVRG following intradiaphragmatic injection.

Figure 2.

WGA identified in the phrenic nuclei in the cervical spinal cord and rostral ventral respiratory groups in the medulla 72 h after intradiaphragmatic injection. (A) Transverse section of the cervical spinal cord displaying anti-WGA positive labeling in the ipsilateral phrenic nuclei (PN, encircled area) following injection of the WGA-HRP-AuNP-proTHP nanoconjugate into the left hemidiaphragm. There is a lack of labeling in the contralateral PN (encircled area). Scale bar is 100 μm. (B) Higher magnification of the ipsilateral PN (encircled area) shown in A displaying fluorescence from the anti-WGA antibody. (C) Higher magnification of the contralateral PN (encircled area) shown in A with a complete lack of anti-WGA label. Scale bar is 50 μm. (D and E) Transverse sections of the medulla at the level of the rostral ventral respiratory groups (rVRGs, encircled area). (D) Ipsilateral rVRG displaying fluorescence from the anti-WGA antibody. (E) Contralateral rVRG displaying fluorescence from the anti-WGA antibody. Scale bar is 100 μm. (F) Higher magnification of the ipsilateral rVRG show in D. (G) Higher magnification of the contralateral rVRG shown in E. Scale bar is 50 μm. P (in A and E) notes pinhole to mark side contralateral to the injection.

Electromyography

Based on the criteria described by Minic et al.,²⁶ results from EMG analysis were used as a Yes/No screening tool to determine which doses should undergo further analysis using the phrenic nerve recording technique. Starting with the lowest tested dose, functional recovery of the diaphragm after a one-time injection of the nanoconjugate dose 0.0005 mg/kg was detected in 66% of the rats (N = 3). Diaphragm activity was detected as early as day 3 and persisted to day 14, the longest time point tested. However, the duration of the burst was less than the non-injured side. Therefore, in each subsequent group, the dose was incrementally increased to find an optimal therapeutic dose (Table 1).

Table 1. Experimental and Control Groups were identified by the solution that was injected into the diaphragm following a C2Hx.

Experimental and Control Groups	Number of rats used for EMG	Number of rats used for bilateral Phrenic Nerve Recording Day 3	Number of rats used for bilateral Phrenic Nerve Recording Day 7	Number of rats used for bilateral Phrenic Nerve Recording Day 14
nanoconjugate 0.0005 mg/kg	3	0	0	0
nanoconjugate 0.0008 mg/kg	4	0	0	0
nanoconjugate 0.0025 mg/kg	4	0	0	0
nanoconjugate 0.0050 mg/kg	7	0	0	0
nanoconjugate 0.0075 mg/kg	7	0	0	0
nanoconjugate 0.01 mg/kg	6	0	0	0
nanoconjugate 0.03 mg/kg	19	8	6	6
nanoconjugate 0.07 mg/kg	25	6	13	12
nanoconjugate 0.12 mg/kg	23	7	9	7
WGA-HRP-AuNP	6	0	5	0
AuNP-proTHP 0.07 mg/kg	8	0	6	0
hydroxyethyl nanoconjugate 0.002 mg/kg	5	0	0	0
hydroxyethyl nanoconjugate 0.3 mg/kg	6	0	0	0

Notes: When possible the number of rats used for bilateral phrenic nerve recordings were also used for EMG to reduce the total number of rats used in the study. Controls: (WGA-HRP-AuNP, AuNP-pro-THP at a dose of 0.07 mg/kg) were assessed for bilateral phrenic nerve recording at day 7. Based on the EMG data the numbers of rats were increased for the 0.03, 0.07, and 0.12 mg/kg doses (see Results).

The doses 0.0008 and 0.0025 mg/kg both resulted in a return of diaphragm activity that occurred in 50% of the rats (N = 4 and N = 4, respectively). Both doses resulted in diaphragm recovery lasting the full 14 days. The doses of 0.005 and 0.0075 mg/kg resulted in the incidence of recovery in only 0.14% of the rats (N = 7 and N = 7, respectively) and the dose 0.01 mg/kg resulted in no observable recovery (N = 6).

When the dose was increased to 0.03 mg/kg recovery was observed in 63% of the rats (N = 19). Similarly, the dose 0.07 mg/kg, resulted in the incidence of recovery as early as day 2 post nanoconjugate injection and lasted up to day 14 in 56% of the rats (N = 25) (Fig. 3), and the 0.12 mg/kg dose resulted in the incidence of recovery in 45% of the rats (N = 20). Interestingly, in 3 rats from the 0.12 mg/kg dose group the LHD burst frequency was minimal but sporadic and sometimes off rhythm compared to the right side, a possible effect of over stimulation.⁶ Thus, we did not go beyond the 0.12 mg/kg dose in this study.

Figure 3.

EMG traces 2 and 14 days post injection 0.7 mg/kg nanoconjugate. Both recordings were obtained from the same rat. (A) 2 days post injection of the dose 0.07 mg/kg nanoconjugate there is a return of the bursting pattern in the left hemidiaphragm (top trace). The bursts in the left hemidiaphragm match those in the right hemidiaphragm (bottom trace). (B) 14 days post injection; the activity has persisted in the left hemidiaphragm (top trace) and remains synchronous with the right hemidiaphragm (bottom trace). In both traces an augmented breath is demonstrated (arrows) followed by a short period of apnea. The augmented breath shows the crossed phrenic pathway is functional.

Injection of a control solution containing a conjugate of just AuNP-proTHP 0.07 mg/kg (no WGA-HRP) (N = 8), or WGA-HRP-AuNP (no proTHP) (N = 6) never produced recovery of the LHD.

Investigation of the hydroxyethyl theophylline nanoconjugate consisted of 2 groups; 0.002 mg/kg N = 5, and 0.3 mg/kg N = 6. The group that received 0.002 mg/kg never produced recovery of the LHD. The group that received 0.3 mg/kg had a 33% occurrence of recovery, however the bursting patterns were sporadic in respect to the pattern produced by the RHD.

Statistical significance of the incidence of functional recovery detected with the EMG analysis was determined using the Chi Square-Fisher’s exact test followed by the Bonferroni adjustment (Fig. 4). Treatment with 0.03, 0.07, or 0.12 mg/kg of the nanoconjugate via intradiaphragmatic injection resulted in a significantly greater proportion of rats achieving functional recovery when compared to the control 0.07 mg/kg AuNP-proTHP; P = 0.003, P = 0.010, P = 0.029 respectively. Treatment with 0.03 mg/kg or 0.07 mg/kg of the nanoconjugate resulted in a significantly greater proportion of rats achieving functional recovery when compared to the dose 0.01 mg/kg; P = 0.015, P = 0.021 respectively. In addition, treatment with 0.03 mg/kg or 0.07 mg/kg of the nanoconjugate resulted in a significantly greater proportion of rats achieving functional recovery when compared to the control WGA-HRP-AuNP; P = 0.015, P = 0.021 respectively. However, with lone EMG data, it is unclear as to whether this is isolated muscle activity stimulated by the injection containing proTHP,^28–33 or if it is a return of function mediated by the bulbospinal pathway via the phrenic nerve.

Figure 4.

Percent of Rats with EMG Recovery of the Left Hemidiaphragm. The chart displays the percentage of rats in each group that had an incidence of recovery detected by EMG of the left hemidiaphragm. Statistical analysis consisted of data from 112 rats. A Chi-square test was performed and significance was found between the proportions of observations (Yes/No) and the groups, X2 (13, N = 112) = 27.075, p = 0.012. Treatment with the 0.03 mg/kg nanoconjugate resulted in significantly greater (#) incidence of recovery compared to 0.01 mg/kg (P = 0.015); WGA-HRP-AuNP (P = 0.015); and 0.07 mg/kg AuNP-proTHP (P = 0.003). Treatment with the 0.07 mg/kg nanoconjugate resulted in significantly greater (†) incidence of recovery compared to 0.01 mg/kg (P = 0.021); WGA-HRP-AuNP (P = 0.021); and 0.07 mg/kg AuNP-proTHP (P = 0.010). Treatment with the 0.12 mg/kg nanoconjugate resulted in significantly greater (‡) incidence of recovery compared to 0.07 mg/kg AuNP-proTHP (P = 0.029).

Phrenic nerve recordings

Based on the screening data obtained from the EMG analysis the doses 0.03, .0.7, and 0.12 mg/kg were selected to analyze the amount of recovery based on the output of the phrenic nerves at days 3, 7, and 14. A representative recording from each day is shown in Fig. 5. The nanoconjugate induced recovery was quantified by analyzing 10 consecutive breaths from bilateral phrenic nerve recordings sampled at days 3, 7, and 14. Controls solutions were sampled on day 7. The area under the curve (AUC) from the right and left integrated (INT) waveforms were calculated and the percent recovery of the left compared to the right was calculated for both the INT and maximal amplitude (MAX). The data used in the following statistical analysis was obtained once physiologically standardized conditions were established.^7,26,34–37 One-way ANOVA with Holm-Sidak pairwise comparison was performed to determine the effect of the nanoconjugate or control solution on the amount of left phrenic nerve (LPN) recovery.

Figure 5.

Day 3 Phrenic Nerve Recordings Analysis. The average percent integrated (AUC) and average percent maximum amplitude (MAX) of the LPN compared to the RPN. Statistical data consisted of neurograms from 18 rats. One way ANOVA showed significant drug effect (#) on day 3 between the 0.12 mg/kg dose group and the remaining two groups, 0.03 mg/kg dose and 0.07 mg/kg dose for the AUC (p < 0.001). No significance was detected from the MAX.

One-way ANOVA showed a significant drug effect on day 3 between the 0.12 mg/kg dose and the remaining two groups, 0.03 mg/kg dose and 0.07 mg/kg dose for the AUC (P < 0.001) but not the MAX (Fig. 5).

One-way ANOVA showed a significant drug effect on day 7 for the AUC between the 0.07 mg/kg dose and the following; 0.03 mg/kg (P < 0.001), 0.12 mg/kg (P < 0.001), WGA-HRP-AuNP (P < 0.001), 0.07 mg/kg AuNP-ProTHP (P < 0.001). There was significant drug effect for the MAX between the 0.07 mg/kg dose and the following; 0.03 mg/kg (P = 0.013), WGA-HRP-AuNP (P = 0.027), 0.07 mg/kg AuNP-ProTHP (P = 0.013) (Fig. 6). Interestingly all control solutions at day 7 resulted in the detectable recovery of the phrenic nerve, however, the level of recovery of the LPN in several groups was negligible.

Figure 6.

Day 7 Phrenic Nerve Recordings Analysis. The average percent integrated (AUC) and average percent maximum amplitude (MAX) of the LPN compared to the RPN. Statistical data consisted of neurograms from 18 rats. One way ANOVA showed significant drug effect on day 7 for the AUC between the 0.07 mg/kg dose and the following; 0.03 mg/kg (p < 0.001), 0.12 mg/kg (p < 0.001), WGA-HRP-AuNP (p < 0.001), 0.07 mg/kg AuNP-ProTHP (p < 0.001). There was significant drug effect for the MAX between the 0.07 mg/kg dose and the following; 0.03 mg/kg (p = 0.013, *), WGA-HRP-AuNP (p = 0.027, **), 0.07 mg/kg AuNP-ProTHP (p = 0.013, ***).

One-way ANOVA showed a significant drug effect on day 14 for the AUC between the 0.03 mg/kg dose and the 0.12 mg/kg dose (P = 0.030). There was also a significant drug effect for the MAX between the 0.03 mg/kg dose and the remaining two doses 0.12 mg/kg (P = 0.001) and 0.07 mg/kg (P = 0.049) (Fig. 7).

Figure 7.

Day 14 Phrenic Nerve Recordings Analysis. The average percent integrated (AUC) and average percent maximum amplitude (MAX) of the LPN compared to the RPN. Statistical data consisted of neurograms from 18 rats. One way ANOVA showed significant drug effect (#) on day 14 for the AUC between the 0.03 mg/kg dose and the 0.12 mg/kg dose (p = 0.030). There was also significant drug effect (†) for the MAX between the 0.03 mg/kg dose and the remaining two doses 0.12 mg/kg (p = 0.001) and 0.07 mg/kg (p = 0.049).

In summary, based on statistical analysis of the neurograms, each dose had a significant drug effect for the three different time points; 0.12 mg/kg on day 3, 0.07 mg/kg on day 7, and 0.03 mg/kg on day 14. There was no single dose tested that produced a consistently significant recovery for all time points compared to the remaining doses. These data may reflect the inconsistent synthesis of the theophylline nanoconjugate (see discussion).

Discussion

Immunostaining

The WGA component of the nanoconjugate was identified by immunostaining in all three rats in the ipsilateral PN in the cervical spinal cord and bilaterally in the rVRGs in the medulla. This shows that the WGA-HRP component of the nanoconjugate was transported to the phrenic nucleus and transsynaptically to the rVRG. However, this technique does not confirm if theophylline is also located in the PN and rVRGs, which must occur for the theophylline to stimulate the CPP. This is due to the hydrolyzable ester bond that enables the detachment of theophylline from the nanoparticle carrier. The primary drug release from ester-based drug linkers on nanomaterials is due to intracellular action of acid hydrolases in the acid compartments such as endosomes (pH 5.0–6.0) and lysosomes (pH 4.8).³⁸ Drug release from ester linkage to nanomaterials tends to be rather slow. For example, only 5% methotrexate was released over a period of 72 h from a dendrimer carrier in a serum medium at 37°C.³⁹ In another study, less than 5% of an antitumor drug SN38, a metabolite of irinotecan, released from its nanocarrier after 24 h in PBS buffer.⁴⁰ In a third study, the ester link of an immunosuppressive agent FK506 conjugated to dextran has a half-life of ∼150 h in a phosphate buffer (pH = 7.4).⁴¹ We are currently conducting in vivo experiments to determine the biodistribution of the nanoconjugate as well as in vitro experiments simulating the in vivo environments to quantify the theophylline release profile from the nanocarrier.

EMG analysis

The results from EMG analysis demonstrate that the nanoconjugate doses 0.0005, 0.0008, 0.0025, 0.03, 0.07, and 0.12 mg/kg stimulate recovery of LHD activity detected as early as day 2 post injection and persisting up to 14 days. Again, based on literature it is likely a significant amount of theophylline is still attached to the nanocarrier at the time of detection of LHD activity. Although current techniques are unable to confirm the presence of theophylline in the PN and rVRGs in this study, we confirmed the presence of WGA by immunostaining techniques. Recovery of the LHD appearing by day 2 post injection corresponds with the time required for the uptake of WGA to the PN and rVRGs.¹⁴

The control solutions, WGA-HRP-AuNP and 0.07 mg/kg AuNP-proTHP did not produce LHD recovery in any of the rats tested. It is interesting to note that AuNP-proTHP was unable to produce LHD recovery even though proTHP was present. The explanation behind this remains unclear; however, without the WGA-HRP there is no transporter to initiate absorptive endocytosis at the phrenic nerve terminals following injection, therefore the drug should not have been delivered to the phrenic nuclei and rVRGs. It is possible for the theophylline to be released in the muscle and have a systemic effect but the dose administered via the nanoconjugate is a fraction of that needed to stimulate CPP recovery following systemic administration.⁶

Bilateral phrenic nerve recordings

To better characterize the effect of the nanoconjugate, quantification of bilateral nerve recordings under standard conditions were compared between groups. The use of ketamine during a procedure to test for respiratory activity may appear problematic. However, in this study all recordings were sampled under the same protocol and there were no issues detecting the activity of the phrenic nerve on the non-injured side. Future, in the study by Giroux et al.,⁴² the effect of ketamine-xylazine anesthesia on respiration in SD rats was described. Rats used in the current study were 10–14 weeks of age (approximately 3 months) selected based on their weight. The study by Giroux et al.,⁴² demonstrates that the concern of respiratory depression following ketamine-xylazine use in younger adult rats (3 months) is only of valid within the first 15–30 min after intraperitoneal administration. At the 45-minute time point Giroux et al.,⁴² found that respiratory frequency had returned to pre-dosing levels in rats 3 months of age.

The variation of the AUC percent recovery observed for the 3 nanoconjugate doses over the course of the study might demonstrate the drug dose effect on respiration as the drug is metabolized and lasting effects over time. For all time points tested the 0.03 mg/kg dose produced the highest percent of LPN activity with an AUC of 2.2%, however, this is modest LPN recovery at best. Fortunately, in addition to theophylline there are multiple drugs known to enhance respiration following high SCI including DPCPX, a specific A₁ adenosine receptor antagonist.^36,43 Simultaneous to this study AuNPs conjugated to WGA-HRP and DPCPX were being investigated.²⁶ Injection of the DPCPX coupled nanoconjugate (0.15 µg/kg DPCPX) produces an average percent recovery of 56.8% on day 7 and 72.4% on day 14 for the AUC of the LPN compared to the right phrenic nerve.²⁶ Further investigation into the use of the DPCPX nanoconjugate is underway.

Challenges in clinical use of the theophylline nanoconjugate

Chemical stability of proTHP

Throughout this work the chemical stability of the theophylline nanoconjugate for in vivo application became apparent. To chemically link theophylline to the AuNPs, a hydroxymethyl group was added to the theophylline creating proTHP, which then forms an ester bond with the capping agent, MSA, on the AuNPs following the Steglich reaction. The synthesized ProTHP was originally stored in vials in ambient air. We discovered that proTHP is reactive to moisture and hydrolyzes back to THP. When comparing 2 nanoconjugate batches (not used in the study), thermogravimetric analysis (TGA) detected a 5.7% weight variance, which is the difference of 105 molecules of proTHP attached to each AuNP. The hydrolysis tendency of proTHP poses significant challenges for the long-term storage of the product for in vivo applications. Following the discovery of the apparent instability of proTHP we developed an experimental protocol to characterize the drug concentration for every batch with TGA to verify uniformity before in vivo use. The integrity of bond stability has been emphasized by others.⁴⁴ It was decided to limit the use of the nanoconjugate for no longer than 30 days post synthesis. To further protect proTHP from hydrolysis we stored the newly synthesized proTHP in argon gas. We suggest that further investigation into the effects of the proTHP-bound nanoconjugate should take proTHP hydrolysis into consideration and carefully quantify the nanoconjugate chemical composition. Interestingly WGA-HRP-AuNP-DPCPX nanoconjugate used in the Minic et al.²⁶ study did not encounter the same issues stated above. In comparison to theophylline, DPCPX is more hydrophobic with an additional 5-carbon ring, a characteristic that may contribute to the chemical stability of pro-DPCPX against hydrolysis.

To increase the stability of the AuNP-proTHP bond, an alternative chemical bond was investigated. The nanoconjugate used in the work described herein used a methyl-binding site; we later identified a version of the proTHP compound that was coupled to the AuNPs using an ethyl-binding site. The change in the binding site resulted in less proTHP molecules bound per AuNP requiring a higher AuNP concentration to maintain the same drug dosage. The effectiveness of the proTHP-coupled nanoconjugate via an ethyl bond was less than the methyl bond. Therefore, efforts to further investigate the ethyl version were abandoned due to low effectiveness.

In vivo stability of the theophylline nanoconjugate

There were notable differences in the AuNP concentration between solutions. To deliver an effective dose of proTHP, the AuNPs in solution were at a concentration of 0.8 mg Au/ml, but at this concentration, we noticed some degree of aggregation of the AuNPs. Nanoparticle aggregation in biological milieu occurs frequently and this limits nanoconjugate uptake by cells (e.g. via endocytosis). The concern of aggregation became evident during an ongoing biodistribution study, while euthanizing rats that received injections 8 weeks prior, the black appearance of the AuNPs, indicative of AuNPs aggregation, was still visible in the diaphragm. If the AuNPs are still visible in the muscle 8 weeks’ post injection one must consider the likelihood of proTHP being delivered to the appropriate nuclei versus the proTHP bond degrading in the muscle. To explore this concern, the concentration of the AuNPs, along with the proTHP, was reduced in a stepwise manner to determine if by lowering the AuNP concentration the probability of AuNP aggregation would be reduced and the probability of nanoconjugate uptake and proTHP delivery would increase. Reduction of the AuNP concentration in solution did result in a reduction of aggregation, as expected, which was confirmed visually by a color change from pinkish red to dark blue, however, the effectiveness of proTHP failed to reach the level of recovery produced by the DPCPX-coupled nanoconjugate. Interestingly the therapeutic dose of DPCPX required only 0.1 mg Au/ml and therefore aggregation complications never arose during the DPCPX study.

Conclusion

WGA-HRP-AuNP-proTHP nanoconjugates are capable of significantly increasing the amount of LPN recovery when injected intradiaphragmatically immediately after C2Hx compared to the control treatments. Targeting delivery of theophylline reduces the therapeutic dose from 15 mg/kg IV⁶ to less than 0.2 mg/kg. In doing so the restored left hemidiaphragm and left phrenic nerve function can be detected as early as 2 days and persists up to day 14 post injection. However, the WGA-HRP-AuNP-proTHP nanoconjugate used in this study fails to induce an equal amount of recovery following the administration of the WGA-HRP-AuNP-DPCPX nanoconjugate.²⁶ The clinical use of theophylline in humans led us to expend maximum effort to find a combination of the proTHP coupled nanoconjugate to reach the effectiveness of the DPCPX coupled nanoconjugate. However, the proTHP nanoconjugate continued to present stability concerns both in terms of its chemical stability in storage and colloidal stability post injection. The amount of proTHP per AuNP for effective recovery requires an AuNP concentration that is prone to aggregation, which limits the transport and cellular uptake of the nanoconjugate. While it is possible to overcome these stability issues by investigating alternative chemical conjugate chemistry, our lab has already yielded a more promising nanoconjugate, DPCPX-containing nanoconjugate.²⁶ The effectiveness of the DPCPX coupled nanoconjugate well-surpassed those of proTHP, therefore all our future efforts will be directed towards the DPCPX-coupled nanoconjugate.

Lastly, the application of the WGA-HRP-AuNP nanoconjugate is not limited to the phrenic motor system. WGA-HRP has been applied over the decades to identify many neuromuscular pathways. In theory, injection of the WGA-HRP-AuNP nanoconjugate bound to a variety of substances could hold promise for investigation and treatment of multiple disease and injury models.

Acknowledgements

The work presented herein is patented: Goshgarian, H. G., Mao, G., Zhang, Y.; Transporter Protein-Coupled Nanodevices for Targeted Drug Delivery. US Patent Publication number: US9649381 B2, May 16, 2017; US20150125926 A1. May 7, 2015. Application number: 14/534,994, Date Filed: November 6, 2014.

Disclaimer statements

Contributors None.

Funding This study was supported by a grant from the Craig H. Neilsen Foundation and National Institutes of Health grant HD-31550 (to HGG). This study was conducted in partial fulfillment of the requirements for the PhD degree at Wayne State University awarded to JLBW.

Conflicts of interest Wayne State University owns the patent therefore; the authors declare no competing financial interests.

Ethics approval None.

ORCID

Janelle Schafer http://orcid.org/0000-0003-1576-6298