Discovery of the Next Generation of Non-peptidomimetic Neurolysin Activators with High Blood-Brain Barrier Permeability: a Pharmacokinetics Study in Healthy and Stroke Animals

Abstract

Purpose

There is growing interest in seeking pharmacological activation of neurolysin (Nln) for stroke treatment. Discovery of central nervous system drugs remains challenging due to the protection of the blood-brain barrier (BBB). The previously reported peptidomimetic Nln activators display unsatisfactory BBB penetration. Herein, we investigate the next generation of non-peptidomimetic Nln activators with high BBB permeability.

Methods

A BBB-mimicking model was used to evaluate their in vitro BBB permeability. Protein binding, metabolic stability, and efflux assays were performed to determine their unbound fraction, half-lives in plasma and brains, and dependence of BBB transporter P-glycoprotein (P-gp). The in vivo pharmacokinetic profiles were elucidated in healthy and stroke mice.

Results

Compounds KS52 and KS73 out of this generation exhibit improved peptidase activity and BBB permeability compared to the endogenous activator and previous peptidomimetic activators. They show reasonable plasma and brain protein binding, improved metabolic stability, and independence of P-gp-mediated efflux. In healthy animals, they rapidly distribute into brains and reach peak levels of 18.69% and 12.10% injected dose (ID)/ml at 10 min. After 4 h, their total brain concentrations remain 7.78 and 12.34 times higher than their A₅₀(minimal concentration required for enhancing 50% peptidase activity). Moreover, the ipsilateral hemispheres of stroke animals show comparable uptake to the corresponding contralateral hemispheres and healthy brains.

Conclusions

This study provides essential details about the pharmacokinetic properties of a new generation of potent non-peptidomimetic Nln activators with high BBB permeability and warrants the future development of these agents as potential neuroprotective pharmaceutics for stroke treatment.

Introduction

Stroke is the fifth leading cause of death in the United States, and more than 80% of cases belong to focal ischemic strokes caused by occlusion of certain cerebral arteries due to thrombus or embolus formation followed by a series of complex pathological cascades and brain injury [1, 2]. Vascular recanalization of impacted arteries within 0 – 24 h of symptom onset through mechanical approaches is effective in a subset of patients [3]. So far, tissue plasminogen activator (tPA) is the only Food and Drug Administration (FDA)-approved drug for ischemic stroke, while the efficacy and safety are compromised if not given to patients within 3 –5 h of symptom onset [4]. There is clearly an unmet clinical need for new pharmaceutics to treat stroke injuries. An increasing number of studies indicate that the administration of neuro protective agents could alleviate the pathological cascades induced by ischemia/reperfusion injury and lessen the size of the infarction and ultimately recovery. Theoretically, recanalization combined with neuroprotection could also maximize brain tissue salvage after stroke onset. Considering the complexity of pathological cascades in stroke, neuroprotective agents acting on multiple molecular and biochemical targets provide a more efficacious therapeutic strategy [5, 6].

Neurolysin (Nln) is a zinc endopeptidase that belongs to the M3 family of metallopeptidases and functions to process various bioactive peptides. Nln widely distributes in the central nervous system (CNS) [7, 8]. Bradykinin, substance P, and neurotensinare among a number of well-known Nln substrates, and their accumulations have been considered to account for stroke progression [9, 10]. Upregulation of Nln is observed in mouse brain and primary cortical neurons after stroke onset [11, 12]. Activation of Nln has been shown as a protective mechanism in experimental stroke models by reducing oxidative stress, excitotoxicity, neuroinflammation, and brain edema [11]. There is a growing interest in seeking pharmacological activators of Nln to protect hypoxic brain tissue from ischemia/reperfusion-induced injury [13, 14].

However, the discovery of CNS drugs remains challenging due to the unique protection afforded to the brain by the blood-brain barrier (BBB), a highly selective, semipermeable border separating the neural tissue from the blood to prevent unwanted plasma substances, including therapeutic agents, to enter the CNS [3]. Therefore, the evaluation of BBB permeability and brain bioavailability is an indispensable step in the discovery of potential stroke therapeutics. We have previously reported the discovery of first-in-class peptidomimetic Nln activators, but they show unsatisfactory BBB penetration in animal studies [15–17]. Herein, we investigated the next generation of non-peptidomimetic Nln activators that possess improved activity, enhanced BBB permeability into the ischemic brain, and high brain bioavailability.

Materials and Methods

Cell Culture

Mouse cerebral endothelial bEnd.3 cells were obtained from the American Type Culture Collection (ATCC) and cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 1% non-essential amino acid solution, and 1% penicillin/streptomycin. MDR1-MDCK cells pre-seeded on 12-well Trasnwell plates were obtained from the MB Biosciences LLC (Cat# ARH T84 12) and cultured following the vendor recommendations. Mouse primary astrocytes were isolated from the cortices of one-day old CD-1 pups (The Charles Rivers Laboratory) as described previously [18]. Briefly, cortices were dissected from the pups’ brains in the Hank’s balanced salt solution (HBSS) supplemented with gentamycin (50 μg/ml). Then, the cortices were digested with 0.25% trypsin for 15 min at 37°C followed by neutralization by complete growth medium (DMEM supplemented with 10% FBS and 1% penicillin/streptomycin). The cells were then seeded into the culture flask with medium change every 3 days until reaching confluency. All cells were maintained in a humidified incubator at 37°C with 5% CO₂.

Fluorometric Enzymatic Assay

The effect of KS compounds on the activity of in-house-produced recombinant rat Nln was carried out essentially as detailed in our recent papers describing discovery of small molecule Nln activators [15, 17].

BBB-Mimicking Model

A BBB-mimicking system with the co-culture of bEnd.3 and astrocytes was established as described previously [19]. Astrocytes were seeded onto the basolateral side of the insert membrane (12-well Corning Transwell insert, 0.4 μm pore size, Lowell, MA) at a density of 150,000 cells per insert, and allowed to adhere for 4 h and grow for another 2 days. Then, bEnd.3 cells were seeded onto the apical side of the same insert membrane at a density of 50,000 cells per insert. The co-cultured cells were allowed to grow for another 7 days to establish barrier integrity with medium change every other day.

In Vitro BBB Permeability Assay

In vitro BBB permeability assays were performed on day 8 with the established co-culture system as described above [16]. The barrier integrity of each Transwell insert seeded with cells was examined by measuring the trans-endothelial cell resistance (TEER) with the EVOM resistance meter and STX-2 electrode (World Precision Instruments, Sarasota, FL). Inserts with TEER value < 80 Ω. cm²were excluded from this assay. To start the assay, the growth medium in both top and bottom chambers was replaced with fresh HBSS buffer, then the test compounds together with sodium fluorescein (NaF) (Sigma Aldrich), a low molecular weight BBB integrity marker, were added to the top chamber at a final concentration of 10 μg/ml. At 30, 60, and 120 min of incubation at 37°C, two 100 μl (total 200 μl) of HBSS buffer were collected from the bottom chambers for LCMS/MS quantitation. The removed volume was supplemented with fresh HBSS buffer to avoid the back diffusion of the tested compounds. The NaF was quantitated by measuring fluorescent absorbance with a microplate reader. The total permeability coefficient $\left({\mathrm{P}\mathrm{e}}_{,t}\right)$ was calculated using the following Eq. (1). dQ/dT is the diffusion rate of tested compounds or NaF crossing the Transwell insert membrane, A is the surface area of the Transwell insert, and ${\mathrm{C}}_0$ is the initial concentration of the tested compounds or NaF added to the top chambers.

$${\mathrm{P}\mathrm{e}}_t=(\mathrm{d}\mathrm{Q}/\mathrm{d}\mathrm{T})\times \left(1/\left[{\mathrm{C}}_0\times \mathrm{A}\right]\right)$$ (1)

Since Transwell insert membrane itself provides some resistance for compounds or NaF movement, blank Transwell inserts without cells (${\mathrm{P}\mathrm{e}}_{\mathit{Blank}}$) were included in the experiments for background subtraction. The Pe values accounted for BBB-mimicking resistance were normalized by the following Eq. (2):

$$\mathrm{P}\mathrm{e}=1/\left(1/{\mathrm{P}\mathrm{e}}_{,t}-1/{\mathrm{P}\mathrm{e}}_{\mathit{Blank}}\right)$$ (2)

Plasma and Brain Protein Binding Assay

The rapid equilibrium dialysis (RED) device (Pierce Biotechnology) was used to test the extent of compounds interacting with mouse plasma and brain proteins according to the vendor manual. Briefly, tested compounds were mixed with mouse plasma or 1:5 diluted brain homogenate to get the final concentration of 500 ng/ml [16]. A 200 μl of each mixture was added to the donor chamber, and 400 μl of Dulbecco’s phosphate-buffered saline (DPBS) buffer was added to the receiver chamber of the dialysis inserts. Then, the dialysis inserts were mounted onto the base plate and incubated at 37°C in an orbital shaker at 250 rpm for 4 h to achieve dialysis equilibrium. At the end of the incubation, an equal volume of solution from the donor and receiver chambers was collected for LCMS/MS quantitation. The portion of unbound fraction $\left(f_{\mathrm{u}}\right)$ was calculated according to Eq. (3) followed by normalization with the dilution factor (D) according to Eq. (4) as described previously [16]. The dilution factors (D) for plasma and brain homogenate samples were 1 and 5, respectively.

$${Measuredf}_{u,}\left(f_{u,m}\right)=\mathrm{R}\mathrm{e}\mathrm{c}\mathrm{e}\mathrm{i}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}/\mathrm{D}\mathrm{o}\mathrm{n}\mathrm{o}\mathrm{r}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}$$ (3)

$$f_u=(1/\mathrm{D})/\left(\left[1/f_{\mathrm{u},m}\right]-1+[1/\mathrm{D}]\right)$$ (4)

In Vitro Metabolic Stability Assay

The in vitro metabolic stability assay in mouse plasma and 1:5 diluted fresh brain homogenate was performed according to previous literature [16]. Briefly, stock solutions of each compound were prepared at 10 μg/ml in LCMS-grade water. Each compound was spiked with equal volume of plasma or brain homogenate to a final concentration of 500 ng/ml in a 1.5 ml tube. The tubes were placed in a water bath at 37°C with slowly shaking for up to 4 h. Sample aliquot (50 μl) was collected at 0, 5, 15, 30, 60, 120, 180,and 240 minutes for LCMS/MS quantitation. The half-life $\left(t_{1/2}\right)$ was estimated as described previously [16].

P-glycoprotein Substrate Assay

MDR1-MDCK cells that overexpress human P-glycoprotein (P-gp) were used to determine if the tested compounds are substrates of P-gp. To perform the experiment, 10 μg/ml of each compound in HBSS was added to either the apical or basolateral compartment of the Transwell inserts in the presence or absence of 10 μM of P-gp inhibitor Cyclosporine A (CsA). For the groups with CsA, the inhibitor was added to both apical and basolateral compartments 30 minutes before the test compounds. After 120 minutes of incubation at 37°C, 100 μl of HBSS buffer were collected twice (total 200 μl) from the opposite compartments (without compound added) for LCMS/MS quantitation. The permeability coefficient (Pe) was calculated according to our previous publication [16]. The Pe values from the apical to basolateral (A > B) direction in the presence and absence of CsA was used to calculate the influx unidirectional flux ratio $\left({\mathrm{U}\mathrm{F}\mathrm{R}}_{A>B}\right)$, and the Pe values from the basolateral to apical (B > A) direction to calculate efflux ${\mathrm{U}\mathrm{F}\mathrm{R}}_{B>A}$ according to Eqs. (5) and (6), respectively. The efflux ratio (ER) was defined by Eq. (7):

$$UF{R}_{A>B}=\left(P{e}_{A>B[+CsA]}\right)/\left(P{e}_{A>B[-CsA]}\right)$$ (5)

$$UF{R}_{B>A}=\left(P{e}_{B>A[+CsA]}\right)/\left(P{e}_{B>A[-CsA]}\right)$$ (6)

$$\mathit{Efflux}\mathit{ratio}\left(ER\right)=\left(P{e}_{B>A}\right)/\left(P{e}_{A>B}\right)$$ (7)

LCMS/MS Analysis

To prepare the samples for LC-MS/MS analysis, each aliquot of His-Tyr, KS52 and KS73 was precipitated by mixing with an organic solvent solution containing (acetonitrile [ACN]: water: formic acid [FA] = 50: 50: 0.1% v/v/v) at ratios of 1:3, 1:4 and 1:9, respectively, followed by thoroughly vortexing for 5 minutes. After that, samples were centrifuged at 14,000 rpm for 10 minutes at 4°C. The supernatants were collected for LCMS/MS quantitation with a system containing a AB SCIEX QTRAP 5500 triple quadrupole mass spectrometer and a Nexera UPLC system (Shimadzu Corporation). RFP column (2.1 mm × 100 mm, 2.7 μm; Raptor FluoroPhenyl, RESTEK, USA) was used following the mobile phases conditions as: (A) water + 0.1% FA, (B) ACN + 0.1% FA; flow rate: 0.4 ml/minute; gradient: 0–0:10 min; 0–10% B, 0:10–4 min; to 90% B, 4 – 5 min; to 10% B.

Pharmacokinetic Experiments in Healthy Mice

Adult male C57BL/6 J mice at 8 – 10 week of age were purchased from the Charles River laboratory, and kept in ventilated cages in a temperature- and humidity-controlled environment along with 12-h of day/night cycles. All procedures were approved by the Institutional Animal Care and Use Committee at the Texas Tech University Health Sciences Center complied with the National Research Council guidelines. For the pharmacokinetic (PK) study, the adult male mice were given 4 mg/kg of each of the tested compounds via jugular vein bolus injection under anesthesia with 1.5% isoflurane mixed with 70% nitrous oxide and 30% oxygen at a flow rate of 1 L/minute. Blood samples (50 μl) were collected via jugular vein at 1, 5, 10, 20, 30, 60, 120, and 240 minutes in 1.5 ml microtubes containing heparin, and centrifuged for 10 min at 6,000 g to obtain plasma. At the last time point, trans-cardiac perfusion was performed with 40 ml of phosphate buffered saline (PBS) (pH 7.4) at a flow rate of 5 ml/minute at room temperature using a Harvard syringe pump. Then, the forebrain without olfactory bulbs, cerebellum, or brain stem was dissected, weighed, and homogenized in water. The plasma and homogenized brain samples were stored at −80°C until LCMS/MS quantitation. The area under the plasma concentration-time curves (${\mathrm{A}\mathrm{U}\mathrm{C}}_{0-\mathrm{T}}$) was determined by using the logarithmic trapezoidal method from the earlier sampling time points where brain concentration did not significantly change. The apparent brain uptake clearance or unidirectional flux constant ($K_{\text{in}}$) was calculated using Eq. (8):

$$K_{\mathrm{i}\mathrm{n}}=Cbr/{\mathrm{A}\mathrm{U}\mathrm{C}}_{0-\mathrm{T}}$$ (8)

$Cbr$ is the brain concentration presented as a percentage of the total injected dose (%ID/ml). $K_{\mathrm{i}\mathrm{n}}$ values were calculated based on total plasma concentration over time, i.e., ${\mathrm{A}\mathrm{U}\mathrm{C}}_{0-\mathrm{T}}$.

Ischemic Stroke Model in Mice

Transit middle cerebral artery occlusion (tMCAO) surgery was performed in male C57BL/6 mice (22 – 24 g) as described previously [16]. Briefly, mice were anesthetized by 1.5% isoflurane mixed with 70% nitrous oxide and 30% oxygen at a flow rate of 1 L/minute. Cerebral blood flow was monitored by a laser Doppler probe (Moor Instruments, Wilmington, DE, USA) that was directly placed over the left middle cerebral artery perfusion region. The animal was placed in the supine position, and a 1.5 cm long incision was made on the midline of the neck. After temporarily occluding the blood flow in the left common carotid artery and left internal carotid artery, a small incision was made in the left external carotid artery and a 6–0 nylon filament with a heat-blunt tip with ~ 0.23 mm diameter was introduced via the incision to the left middle cerebral artery bifurcation. An 80% or more blood flow reduction was considered successful occlusion. After 1 h, the filament was removed to allow blood perfusion restoration. A 70% or more blood flow increase was considered successful reperfusion. Animals without satisfied occlusion or reperfusion were excluded. The 2,3,5-triphenyl tetrazolium chloride (TTC) staining was randomly performed on some animals to confirm the successful induction of ischemic stroke [16].

Pharmacokinetic Experiments in Stroke Mice

To compare the PK profile of KS52 and KS73 in healthy and stroke animals, 4 mg/kg of each compound was given to adult male C57BL/6 J mice after1 hour of tMCAO and 3 h of reperfusion via jugular vein as described previously [16]. Blood samples were collected at 1, 5, 10, 20, and 30 min. The forebrain was extracted after trans-cardiac perfusion and separated into ipsilateral (left) and contralateral (right) hemispheres. Plasma and brain samples were processed for LCMS/MS analysis. Pharmacokinetic profile was analyzed as aforementioned description.

Statistical Analysis

Statistical analysis was performed with Prism 9.5.0 software (GraphPad Software). Data were presented as mean ± SEM. Data were analyzed by one-way analysis of variance (ANOVA) followed by a Tukey’s post hoc multiple comparison test. p value < 0.05 was considered as statistically significant.

Results

In Vitro BBB Permeability, Plasma and Brain Protein Binding, and Metabolic Stability

The design and synthesis of the new non-peptidomimetic Nln activators (Fig. 1A) are described in a recent U.S. Patent Application (Application Number 63/501,273) [20]. Their peptidase activity was measured as described previously [15]. According to the data, the A₅₀ (minimal concentration required for enhancing 50% peptidase activity) of KS52 was 0.96 μM and that of KS73 was 0.74 μM, which were almost one order lower compared to that of the previous peptidomimetic activator 11a (7.0 μM) (Table I), suggesting a markable improvement in potency.

Fig. 1.

Apical to basolateral transport of non-peptidomimetic Nln activators through a BBB-mimicking model. (A) Chemical structure of non-peptidomimetic Nln activators. (B) In vitro permeability coefficient (Pe) of compounds KS52 and KS73 compared to the endogenous activator His-Tyr. HY: His-Tyr; cm/sec: cm²/second; Data were shown as Mean ± SEM, n = 3. *: p < 0.05, **: p < 0.01, ns: no statistical significance.

Table I.

In Vitro Protein Binding and Metabolic Stability in Mouse Plasma and Brain

	A50 (μM)	$f_{\mathrm{u}}$ plasma (%)	$f_{\mathrm{u}}$ brain (%)	$t_{1/2}$ plasma (minute)	$t_{1/2}$ brain (minute)
KS52	0.96	9.77 ± 0.27	0.17 ± 0.02	> 1000	567.37 ± 165.45
KS73	0.74	27.74 ± 1.16	1.95 ± 0.31	> 1000	825.00 ± 269.44
11a	7.0	41 ± 8.1	62 ± 0.01	248.25 ± 28.1	181.55 ± 14.9 (15)

$f_{\mathrm{u}}$, unbound fraction; $t_{1/2}$, half-life. The assays were done in triplicate. Data were shown as Mean ± SEM

To examine the in vitro BBB permeability, each of the tested compounds was added to the luminal compartment of an established co-culture system and the progressive penetration through the BBB model was measured. As shown in Fig. 1B, a dramatic increase in the apical to basolateral (A > B) transport for the new, non-peptidomimetic activators (KS52 and KS73) was observed. The permeability coefficient (Pe) for these compounds was 47.9- and 82.7-fold higher than that of the dipeptide activator His-Try (5.75 ± 0.94 × 10⁻⁵ and 9.92 ± 1.70 × 10⁻⁵ vs. 0.12 ± 0.02 × 1 0⁻⁵ cm²/second, p < 0.05). Furthermore, these Pe values were also 3.0 – 5.2 times higher than that of the peptidomimetic activator 11a (1.90 × 10⁻⁵ cm²/second) from previous studies [15]. The Pe values for the BBB integrity marker, NaF, were not significantly different among the three groups (data not shown), ensuring that BBB-mimicking integrity was maintained.

Plasma and brain protein binding of these compounds was analyzed to determine the percentage of unbound fractions $\left(f_{\mathrm{u}}\right)$ in plasma and brain homogenate. Compound KS73 showed relatively higher $f_{\mathrm{u}}$ in mouse plasma and brain homogenate (27.74 ± 1.16% and 1.95 ± 0.31%, respectively) compared to compound KS52 (9.77 ± 0.27% and 0.17 ± 0.02%, respectively) (Table I). The in vitro half-life $\left(t_{1/2}\right)$, defined as the time needed for 50% degradation of the tested compound, was estimated according to assumed pseudo-first-order degradation. Both KS52 and KS73 showed comparable but increased stability in plasma (> 1000 minutes vs. 248.25 ± 28.1 minutes) and improved brain stability (567.37 ± 165.45 and 825.00 ± 269.44 minutes vs. 181.55 ± 14.9 minutes, respectively) compared to the peptidomimetic activator 11a (Table I) [15].

KS52 and KS73 are not Substrates of P-gp Transporter

As an important efflux transporter in the BBB, P-gp plays a crucial role in preventing unwanted plasma substance to enter the CNS. Here we used P-gp-overexpressing cells that are transfected with human MDR1 gene to examine the P-gp-mediated efflux in BBB [21]. Unidirectional flux ratio (UFR) together with efflux ratio (ER) can predict if an agent is a substrate of P-gp. The UFR and ER values of < 2 suggest one agent is not a substrate of P-gp [21, 22]. As shown in Fig. 2A–C, the Pe values for these compounds in the presence of CsA were not significantly different from that in the absence of CsA, in either influx or efflux direction. Both compounds’ Pe values in the efflux (B > A) direction was not different from that in the influx (A > B) direction. A summary of UFR and ER values are shown in Table II. Collectively, the results clearly indicate that these non-peptidomimetic activators are not substrates of P-gp.

Fig. 2.

Permeability of KS52 and KS73 through MDR1-MDCA monolayer. (A, B) Permeability coefficient (Pe) of KS52 and KS73 in the presence and absence of P-gp inhibitor CsA, in either apical to basolateral (A > B) (A) or basolateral to apical (B > A) (B) direction. (C) Comparison of influx (A > B direction) and efflux (B > A direction) mediated by P-gp. CsA: cyclosporine A; cm/sec: cm²/second; Data were shown as Mean ± SEM, n = 3. ns: no statistical significance.

Table II.

Summary of UFR and ER Values

	UFR (A > B)	UFR (B > A)	ER
KS52	1.06 ± 0.01	0.97 ± 0.14	1.02 ± 0.16
KS73	1.11 ± 0.10	1.04 ± 0.08	0.85 ± 0.09

UFR, unidirectional flux ratio: ER, efflux ratio: Pe, permeability coefficient. The assays were done in triplicate. Data were shown as Mean ± SEM

Pharmacokinetic Profile of Single Dosing of KS52 and KS73 in Healthy Mice

To elucidate the pharmacokinetic (PK) profiles, 4 mg/kg of each compound was administrated to C57BL/6 mice through intravenous bolus injection. Plasma samples were collected at different time points, then the forebrain was dissected. As depicted in Fig. 3A, the plasma profiles for both compounds showed a biexponential decline over a 4-hour sampling period. KS73 exhibited relatively higher plasma concentrations than KS52 at each time point. A biexponential decline in the brain for both compounds was observed over the sampling period (Fig. 3B). The data show that these new compounds rapidly distributed into the brain and reached peak levels of 18.68 ± 0.20 and 12.10 ± 0.62 %ID/ml at 10 minutes. This translates to total brain concentrations of KS52 and KS73 being ~ 53.60 ± 4.19 and 35.82 ± 1.11 μM (Fig. 3B, C), respectively, which is 55.83- and 48.40-fold higher than their respective A₅₀value. At the longest time point tested in this study, brain uptake for these compounds remained at 2.69 ± 1.33 and 2.65 ± 0.33 %ID/ml which equals to 7.46 ± 0.72 and 8.23 ± 1.16 μM, respectively. Notably, the total brain concentrations at the last time point were still 7.79- and 11.12-fold higher than their respective A₅₀ values (Fig. 3C). Furthermore, apparent $K_{\text{in}}$ value was calculated based on the brain concentration (%ID/ml) over plasma ${\mathrm{A}\mathrm{U}\mathrm{C}}_{0-\mathrm{T}}$ for the earlier time points where brain concentrations did not significantly decrease (Fig. 3D). As shown in Fig. 3C, the brain concentrations for both compounds did not significant decrease within the first 30 minutes (p > 0.05, data not shown), therefore, the apparent $K_{\text{in}}$ values within 30 minutes were calculated with Eq. (8). According to the data, the apparent $K_{\text{in}}$ values for KS52 and KS73 were ranged from 7931.67 ± 2589.83 to 2425.67 ± 192.56 μl.min⁻¹.g⁻¹ and from 3143.64 ± 395.15 to 2298.09 ± 25.19 μl.min⁻¹.g⁻¹, respectively. It is worthy of note that the brain uptake of the previously reported peptidomimetic Nln activator (11a) was only around 0.04 ± 0.005% ID/ml and its apparent $K_{\text{in}}$ values ranged from 12.31 ± 1.86 to 5.01 ± 1.2 μl.min⁻¹.g⁻¹ during the sampling time points [16]. Taken together, these non-peptidomimetic activators possess enhanced brain uptake compared to the peptidomimetic activators indicated by several folds higher $K_{\text{in}}$ and total brain concentrations.

Fig. 3.

Pharmacokinetic profile of KS52 and KS73 after intravenous bolus injection in healthy mice. (A) Concentration-time profiles of KS52 an KS73 showed biexponential decline in plasma. Left panel showed the absolutive values, right panel showed the relative portion to injection dose. (B) Concentration-time profiles of KS52 an KS73 showed biexponential decline in brains. Left panel showed the absolutive values, right panel showed the relative portion to injection dose. (C) Brain concentrations relative to A₅₀ over multiple time points. (D) Brain uptake clearance kinetics ($K_{\text{in}}$) by using multiple time point (up to 30 min) analysis. Cplasma and Cbr: concentration in plasma and brain; %ID/ml: percentage of injection dose/ml. A₅₀: minimal concentration required for enhancing 50% peptidase activity. Data are presented as Mean ± SEM, n = 3. ***: p < 0.001, ****: p < 0.0001. ns: no statistical significance.

Plasma Pharmacokinetic Profile of KS73 in Stroke Mice

To investigate the PK profile in stroke animals and compare it with healthy animals, tMCAO surgery was performed on C57BL/6 mice following the 1-h occlusion and 3-h reperfusion procedure similar to previous studies [11]. Since both compounds showed a similar PK pattern in healthy animals, KS73 was randomly chosen to be tested in the stroke model. Animals with successful occlusion and reperfusion were included in this study (Fig. 4A). According to the data, the plasma concentration-time profile of KS73 showed a biexponential decline in both stroke and healthy groups, with relatively higher levels in the stroke group (Fig. 4B). Consistently, the ${\mathrm{A}\mathrm{U}\mathrm{C}}_{0-30\text{minutes}}$ in stroke group was about 1.63-fold higher than that in the control group (6.50 ± 0.38 vs. 3.99 ± 0.23% ID.ml⁻¹.min, p < 0.001) (Fig. 4C). The increased plasma ${\mathrm{A}\mathrm{U}\mathrm{C}}_{0-30\text{minutes}}$ was also observed with the peptidomimetic activator 11a in the previous study [16].

Fig. 4.

Plasma pharmacokinetic profile of KS73 in stroke mice. (A) Representative TTC staining of brain slices after 1-h tMCAO-induced occlusion and 3-h reperfusion (Left panel) and cerebral blood flow measured by a laser Doppler probe presented as percentages of the baseline. (B) Plasma profile in stroke and healthy animals. Left panel showed the absolutive values, right panel showed the relative portion to total injection dose. (C) The area under the plasma concentration-time curves $\left({\mathrm{A}\mathrm{U}\mathrm{C}}_{0-30\mathrm{m}\mathrm{i}\mathrm{n}}\right)$ in stroke (tMCAO) and healthy (Control) mice. TTC: 2,3,5-triphenyl tetrazolium chloride. %ID/ml: percentage of injection dose/ml. tMCAO and Control: transit middle cerebral artery occlusion animals and healthy animals. Data were presented as Mean ± SEM, n = 4. **: p < 0.01, ***: p < 0.001, ns: no statistical significance.

Brain Uptake of KS73 in Stroke Mice

Single time point brain uptake for KS73 in stroke mice was quantitated as aforementioned in healthy mice. After 30 minutes of dosing, the brain uptake was not significantly different among healthy brains, ipsilateral hemispheres and contralateral hemispheres of stroke mice (10.32 ± 0.46 vs.10.48 ± 0.63 and 11.22 ± 0.76% ID/ml, p > 0.05) (Fig. 5A). The apparent $K_{\text{in}}$ values of ipsilateral and contralateral hemispheres were comparable (396.07 ± 14.21 and 426.88 ± 33.65 μl.min⁻¹.g⁻¹, p > 0.05) at 30 minutes, but were lower than that of healthy brains (863.71 ± 31.28 μl.min⁻¹.g⁻¹) (p < 0.001) (Fig. 5B). According to Eq. (8) used in this study, the lower apparent $K_{\text{in}}$ values in stroke brains might be due to the aforementioned relatively higher plasma ${\mathrm{A}\mathrm{U}\mathrm{C}}_{0-30\text{minutes}}$ in stroke animals instead of absolutive brain concentrations, i.e., increased plasma exposure lowers the brain uptake clearance when the brain concentration did not significant change. Finally, all ipsilateral hemispheres, contralateral hemispheres, and healthy brains exhibited comparable absolutive brain concentrations (36.28 ± 1.26, 39.12 ± 3.09 and 39.26 ± 1.71 μM), which were ~ 50-fold higher than the A₅₀ of KS73 (Fig. 5C).

Fig. 5.

Brain uptake of KS73 at a single time point analysis in stroke mice. (A) Brain uptake among stroke ipsilateral hemisphere, stroke contralateral hemisphere, and healthy brain. Left panel showed the absolutive values, right panel showed the relative portion to total injection dose. (B) Brain uptake clearance kinetics ($K_{\text{in}}$). (C) Achieved brain concentration relative to A₅₀. Cbr: concentration in brain; %ID/ml: percentage of injection dose/ml; tMCAO-Ips, tMCAO-Con and Control: ipsilateral hemisphere of stroke animal, contralateral hemisphere of stroke animal, and healthy brains. A₅₀: minimal concentration required for enhancing 50% peptidase activity. Data are presented as Mean ± SEM, n = 4. ****: p < 0.0001, ns: no statistical significance.

Discussion

Discovery of new CNS drugs presents consistently more challenges due to the unique protection afforded to the brain by the BBB, which is a highly selective, semipermeable border separating the neural tissue from the blood and restricting the movement of larger molecules and unwanted plasma substances [3]. In this study, we report a new class of non-peptidomimetic Nln activators with markable improvement in bioactivity and significantly increased BBB permeability and brain bioavailability. These new activators show improved half-lives yet high protein binding in the plasma and brain homogenate compared to the previous peptidomimetic activators, suggesting increased metabolic stability but decreased unbound fraction in the plasma and brain. These non-peptidomimetic activators exhibit a biexponential decline in plasma and brains shown as a rapid distribution phase followed by elimination phase. Following single dosing of 4 mg/kg via intravenous bolus injection, they rapidly enter the brain of healthy animals within minutes. Approximately 18.69% and 12.10% of administration dose is detected at 10 minutes, which is ~ 300 – 400 times higher than that of the peptidomimetic activator 11a [15]. Total brain concentrations range from 36.60 – 7.47 μM and 35.82 – 9.13 μM over the sampling period for each compound, which are also markedly higher than that of 11a. Considering the increased activity indicated by the sub-micromolar A₅₀, the achieved total brain concentrations suggest these compounds might activate Nln in the healthy and ischemic brains for multiple hours. However, it worthy of note that the unbound fraction ($f\mathrm{u}$) of these compounds are relatively low in the brain, indicating that the unbound concentrations, which is more associated with pharmacological efficacy, might be low in the brain. For example, single dose of 4 mg/kg of KS73 reached 35.82 ± 1.11 μM of total brain concentration which could be converted to 0.70 μM of unbound concentration after corrected by the $f\mathrm{u}$ value (~ 2%), which is slightly lower than its A₅₀ (0.74 μM). Same dose of KS52 may result in even lower unbound concentration in the brain. It could be expected that increased and/or repeated doses of KS73 (but not KS53) might achieve higher unbound concentrations in the brain than its A₅₀. Furthermore, measuring the unbound brain-to-plasma partition coefficient ($K_{\text{p,uu,}\text{brain}}$), a parameter with a direct quantitative linking to the therapeutic dose, will also be helpful for the insight of the unbound concentrations of these highly protein binding compounds and should be included in future studies, along with structural modifications that could reduce protein binding. Furthermore, as shown in our previous publication, after intravenous dosing, the peptidomimetic activator 11a exhibits a central (plasma) distribution pattern, i.e., plasma levels are much higher than brain levels over the sampling period [16]. The current study reveals that the non-peptidomimetic activators have plasma levels that are dramatically lower than brain levels over the similar sampling period. This pattern might reflect the rapid penetration of a large portion of the non-peptidomimetic activators into the peripheral tissues including the brain, which results in only a small portion remaining in plasma.

Besides the normal physiological conditions, we also investigated the PK profile in stroke animals and compare it with control animals. Pathophysiologic PK analysis is more relevant to clinical conditions and is more precise to predict the disposition of a therapeutic agent. In the early phase of acute ischemic stroke, ischemic core is formed due to the complete blockage of blood perfusion. In the case of delayed reperfusion, the ischemic core is highly voluntary to develop infarction that is surrounded by reversibly injured and salvageable tissues known as penumbra. The survival of neurons in the penumbra is especially important for the stroke outcome [23–25], which could be salvaged by early vascular recanalization and proper penetration and distribution of neuroprotective reagents [22, 25]. Our data suggest that non-peptidomimetic activators can rapidly distribute into ischemic hemispheres, as they do in the healthy animals, suggesting they could efficiently reach the penumbra under stroke condition, too.

It has been shown that from 3 to 48 h after occlusion, reperfusion results in reperfusion-induced BBB damage, hence brain edema and additional brain injuries [26]. This BBB damage may increase the brain entry of low-BBB-penetrative agents, such as the peptidomimetic activator 11a, via opening BBB and/or altered influx and efflux transporter [16]. However, we do not observe a similarly increased brain uptake of the non-peptidomimetic activators in the stroke hemisphere. It might be due to the fact that the physiologic BBB permeability is high enough to mask the impact of opening BBB induced by ischemia. Another possible reason is that these compounds might be less affected by altered transporter expression at the ischemic BBB. It has been reported that certain BBB transporters could be upregulated in stroke. Previous studies, including our lab, have shown that organic anionic transporter 1 (OATP1) and organic cation transporter1 (OCT1) are selectively upregulated and contribute to the improved BBB penetration of an opioid receptor agonist, biphalin, diabetes mellitus medicine, metformin, and N-methyl-D-aspartate (NMDA) receptor antagonist, memantine, in cerebral endothelial cells and brain microvessels exposed to ischemia [27–29]. An alternative hypothesis is that the upregulation of any putative influx transporters or downregulation of efflux transporters, if any existed, compensates for each other. Future studies are needed to address these questions. Furthermore, similar to that of the peptidomimetic activator 11a [16], we observe an increased plasma ${\mathrm{A}\mathrm{U}\mathrm{C}}_{0-30\text{minutes}}$ of non-peptidomimetic activators in stroke animals, although the underlying mechanisms are not clear.

Another difference between the two generations of Nln activators is in the ability to serve assubstrates of P-gp. In drug development and research, P-gp is one of the major focuses due to its broad substrate specificity and abundant expression in the BBB. P-gp is an ATP dependent efflux protein mainly exists in the apical compartment of BBB and functions to actively back transport various molecules to the blood [30]. Although the peptidomimetic activators are substrates of P-gp [16], the non-peptidomimetic activators do not serve as substrates of P-gp. This might help to partially explain the different BBB permeability between these structures. Besides P-gp, Breast Cancer Resistance Protein (BCRP), which can synergize with P-gp, and Multidrug Resistance-Associated Proteins (MRPs) are two additional BBB efflux transporters that can significantly affect the overall brain uptake of drugs [31, 32]. Future studies are warranted to study the affinity of the non-peptidomimetic activators to BCRP and MRPs to verify if they are substrates of these efflux transporters or not.

This preliminary study elucidates the PK profile and brain bioavailability of a single intravenous bolus dose. The blood-brain barrier transport mechanisms and therapeutic efficacy of these compounds are still not clear. Further work is required to understand the PK profile of multiple dosing, the role of other transporters in mediating the influx and efflux transportation in the CNS, and the therapeutic efficacy of these new activators.

Conclusion

In the present study, we report about pharmacokinetic properties of the next generation of non-peptidomimetic Nln activators that process enhanced activity, improved stability and high BBB permeability leading to optimal brain bioavailability in healthy and stroke animals. This pilot pharmaceutical study warrants the future development of this new class of Nln activators as potential neuroprotective agents for stroke therapy.

Data Availability

The authors confirm that the data supporting the findings of this study are available within the article.