Intranasal delivery of darunavir improves brain drug concentrations in mice for effective HIV treatment

Abstract

Despite the availability of combined antiretroviral therapy (cART), which reduces the HIV replication in chronically HIV-infected patients, HIV associated neurocognitive disorders (HAND) persists in the brain. The blood-brain barrier (BBB) is the major barrier for the penetration of drugs including antiretrovirals, limiting the drug penetration to the brain. In the present study, we have shown improved brain drug concentration in mice for darunavir (DRV), an FDA-approved drug, using an intranasal (IN) delivery method that bypasses the BBB. Here, we compared the time-dependent biodistribution of DRV at two different concentrations, high (25 mg/kg) and low (2.5 mg/kg), using two administration routes intravenous (IV) and intranasal (IN) in brain, liver, lungs, and plasma. Compared with IV administration, IN administration demonstrated a significantly improved DRV penetration in the brain at both low and high DRV concentrations (IV vs IN: at 2.5 mg/kg: 6.91 ± 1.69 ng/g vs 12.08 ± 2.91 ng/g, at 25 mg/kg: 12.84 ± 2.88 ng/g vs 19.74 ± 1.80 ng/g). As expected, IN administration showed significantly lower DRV concentrations in plasma (IV vs IN: at 2.5 mg/kg: 81.37 ± 22.04 ng/g vs 19.91 ± 12.65 ng/g, at 25 mg/kg: 899.12 ± 136.93 ng/g vs 320.56 ± 40.04 ng/g) and liver (IV vs IN: at 2.5 mg/kg: 118.39 ± 28.13 ng/g vs 29.27 ± 4.17 ng/g at 25 mg/kg: 1085.18 ± 255.0 ng/g vs 833.83 ± 242.4 ng/g). The IN administration did not show significant change in lungs compared to the IV administration. As a result, these findings suggest that the IN route can increase the DRV level in the brain, suppressing HIV in the brain reservoirs. Additionally, it could also reduce off-target effects, especially in peripheral organs.

Highlights

• •Antiretrovirals show suboptimal concentrations of the drugs in the brain, which leads to persistent HIV replication.
• •Brain distribution of darunavir using intranasal and intravenous administrations was studied in a mice model.
• •Compared with intravenous, intranasal administration showed significantly improved drug concentrations in the brain.
• •Compared to intravenous, intranasal administration showed significantly decreased drug concentrations in plasma and liver.
• •Findings from this study suggest that intranasal route could be effective in treating HIV in the brain reservoirs.

1. Introduction

HIV treatment involves the use of antiretroviral drugs that reduces the systemic HIV viral load and mortality in people living with HIV (PLWH). Globally, more than 16 million cases of HIV-associated neurocognitive disorders (HAND) have been reported in HIV-infected adults, with 72% prevalence of HAND in HIV-infected adults in sub-Saharan Africa [1]. Although cART has transformed HIV to a chronic infection with improved life expectancy, issues related to inflammatory response, drug toxicity, and aging associated neurodegeneration remain persist and responsible for HAND [2,3].

Darunavir (DRV) (a class of protease inhibitors) along with nucleoside reverse-transcriptase inhibitors (NRTIs) is WHO recommended second-line regimen for treating and preventing HIV infection [4]. In a two-by-two factorial, open-label, noninferiority trial study, DRV-based second line regimen showed effective viral suppression in the population of patients, including patients for whom NRTIs are predicted to have little or no activity [4]. However, DRV show limited HIV suppression in the brain, even with the use of well tolerated combination antiretroviral therapy (cART) [5,6]. A comparative study that involved HIV positive and HIV negative participants from the pre-cART era and cART era showed significant impairment in motor skills, cognitive speed, and verbal fluency, whereas HIV positive patients during cART era showed more memory (learning) and executive function impairment [7]. Hence, HIV can persist in the brain tissue of PLWH and develop cognitive and motor impairments collectively known as HAND during the cART era [6,8,9].

The drug's permeability to reach the brain is mostly hampered by the blood-brain barrier (BBB) [10]. The endothelial cells of the brain capillaries that constitute the blood-brain barrier (BBB) express multiple solute carriers and ATP binding cassette (ABC)-type efflux transporters [11]. These efflux transporters are transmembrane proteins responsible for the permeability and efflux of endogenous compounds and xenobiotics [12] from the brain parenchyma. BBB is reported to have various efflux transporters including P-glycoprotein (P-gp), breast cancer resistance protein (BCRP), and multi-drug resistance Protein (MRP) isoforms [[13], [14], [15]]. These efflux transporters also limit the entry of protease inhibitors such as DRV into the brain. However, the difference in drugs' inhibitory effects on these efflux transporters may have an impact on the permeability of cross the BBB [16]. DRV efflux by drug transporter P-gp can be reduced in the presence of increasing concentrations of ritonavir (RTV), suggesting a dose-dependent inhibition of transporter efflux activity [16,17]. RTV also increases DRV concentration by inhibiting DRV-metabolizing cytochrome P450 3A4 (CYP3A4) enzyme [[18], [19], [20], [21]]. However, combining the drugs often causes drug-drug interactions (DDI) and drug-induced toxicity, especially in the brain [20,22]. In addition to the presence of efflux transporters on BBB, expression of efflux transporters and CYP3A4 on glial cells such as microglia and astrocytes suggests that glial cells may also play a role in the suboptimal drug concentration in the brain [23,24].

It is crucial to develop a therapeutic strategy for drug delivery to the brain that can avoid first pass liver metabolism and bypass the BBB.W. H. Frey II first developed the non-invasive intranasal (IN) method for bypassing the BBB to deliver neurological therapeutic agents to the brain along the olfactory neural pathway [25,26]. IN administration delivers therapeutics to the central nervous system primarily through two pathways, one is by traveling along the olfactory axon bundles located on the olfactory epithelium at the roof of the nasal cavity. This allows therapeutics to bypass the BBB to reach the olfactory bulb and rostral brain regions. The other pathway is via the trigeminal neural pathway by following the ophthalmic and maxillary branches from nasal mucosa to reach the brainstem and spinal cord [[27], [28], [29]]. Lochhead et al. used an insulin formulation to determine the distribution after IN administration. Intranasal insulin is to bypass the BBB to reach the widely distributed throughout cerebral perivascular spaces. Intranasal therapeutics have shown benefit in animal models including reducing stroke damage, reversing Alzheimer's neurodegeneration, improving memory, stimulating cerebral neurogenesis, and treating brain tumors [29]. The IN administration of insulin also improved learning and memory of healthy humans and Alzheimer's disease patients. A recent study also showed that IN insulin can improve the verbal and visuospatial memory and cognition function in type 2 diabetic patients [30]. The potential application of insulin through IN administration is also under clinical trials for Parkinson's disease, cognitive Impairment, hyperlipidemia, hypoglycemia unawareness, etc. (clinicaltrials.gov) In this study, we used IN routes to determine its ability to enhance the delivery of DRV in the brain compared to IV route.

2. Material and methods

2.1. Materials

DRV (D193500) was purchased from Toronto Research Chemicals, Inc. (Ontario, Canada). Sterile phosphate-buffered saline (PBS) (10100-031) was obtained from Gibco (Dublin, Ireland). LC/MS-grade acetonitrile (A955) and formic acid (AC270480010), BD PrecisionGlide 25G needle (14-826-49), and BD 1 Ml TB syringe (14-826-88) were obtained from Fisher Scientific (Hampton, NH, USA.

2.2. DRV administration via intravenous (IV) and intranasal (IN) routes in mice

Ten-twelve weeks old male and female Balb/c mice were purchased from Jackson Laboratory (Bar Harbor, MA) and were acclimated to the animal facility for at least 7 days. Five mice per cage were housed in a sterile room with 12/12-h light-dark cycles. Temperature and humidity were maintained at a constant level in the room. There was free access to food and water. Animal studies were performed according to The University of Tennessee Health Science Center Institutional Animal Care and Use Committee (UTHSC-IACUC) protocol. For a DRV study, a total of 48 mice were randomly divided into two groups (for IV and IN) and four subgroups for four-time points (1, 3, 6, and 12h, 3 females and 3 males in each group). Mice were administered using IV and IN routes with 2.5 mg/kg and 25 mg/kg of DRV dissolved in 5% DMSO, 80% PEG400, and 15% PBS. The concentration of DRV in the final solution was prepared and adjusted to 0.5 ml/kg for IN administration in mice. The final DMSO volume was 0.025 ml/kg, which is lower than the nontoxic dose used in previous study [31]. Relatively high (25 mg/kg) drug concentration of DRV is comparable to that of EVG used previously, which is non-toxic and physiologically relevant for animal studies [23,24]. We also used 2.5 mg/kg, mainly because high DRV concentration via IN route may not be feasible, especially when formulated in nanoparticles, as well as high DRV concentration may show neurotoxicity in long-term treatment. Mice were euthanized by exsanguination under deep isoflurane anesthesia. Blood and tissues (brain, liver, and lungs) were collected at time points 1, 3, 6, and 12 h. Blood was obtained using cardiac puncture and collected in EDTA-containing blood-collection tubes. Mouse blood was settled down at room temperature and then centrifuged at 6000 rpm, for 10 min at 4 °C, for plasma harvesting. Mice plasma and brains were placed into tubes and frozen at −80 °C, until further analysis by LC-MS/MS. Tissue samples were homogenized in 1XPBS (1:4 [wt/vol]). Fifty μl of each sample was used for LC/MS-MS. All experimental protocols involving the use of laboratory animals were approved by the UTHSC Institutional Animal Care and Use Committee (IACUC).

2.3. DRV quantification using LC-MS/MS

Mouse plasma and tissue samples were analyzed for DRV concentration, using our standardized LC-MS/MS method as previously described [18,32]. Briefly, DRV and internal standard (RTV, 50 ng/ml) were quantified using a tandem mass spectrometer AB SCIEX Triple Quad 5500 that was equipped with electron spray ionization in positive mode. Compounds were isolated by a liquid chromatographic system (LC-20AD XR from Shimadzu, MD). The acquisition was performed in multiple reaction monitoring (MRM) mode, and the MultiQuant® software (AB Sciex, Foster City, CA) was used for data acquisition and processing. To reduce matrix effects, calibration curves were prepared with plasma or tissue homogenates based on the sample types. No significant interference from the tissue and plasma samples were detected. The linear calibration curves range from 1 to 500 ng/ml with a correlation coefficient (r²) of 0.995 and a weighting factor of 1/x². The lower limit of quantification (LLOQ) was 1 ng/ml. The observed retention time of DRV was 2.25 min. DRV standards and experimental samples were extracted by adding 4 vol (200 μl) of cold acetonitrile for plasma, livers, lungs, and methanol for brains, which contained 50 ng/mL RTV as internal standard. The mixed solution was vortexed, centrifuged, and analyzed using the LC-MS/MS. The optimized mass and chromatographic settings were listed in Table 1 referring to calibration standards and methods validation. The LCMS spectra and multiple reaction monitoring (MRM) chromatogram for DRV is presented in Supplement Figure 1.

Table 1.

Bioanalytical method summary for the analyte DRV and the internal standard RTV.

Optimized mass parameters
Analyte	Transitions (m/z)	CE (eV)	CXP (eV)
Darunavir (DRV)	547.7/392	20	30
Ritonavir	721.3/296.1	27	29
Other parameters: DP: 100, EP: 10, CUR: 20, CAD: 8, IS: 5500, GS1: 50, GS2: 50, Temp: 500
Liquid chromatography conditions
HPLC Column: Waters Xterra® MS C18 Column, 125 Å, 3.5 μm,4.6 mm × 50 mm
Mobile Phase: acetonitrile and water with 0.1% formic acid Injection volume: 1 μl

Total flow: 1 ml/min
Gradient Time	Organic solvent (%)
0.10	30
1.00	50
1.50	60
5.00	60
5.10	30

MS parameters are abbreviated as follows: Capillary electrophoresis (CE), Cell exit potential (CXP), Declustering potential (DP), Entrance potential (EP), Curtain gas (CUR), Collision gas (CAD), Ion spray velocity (IS), Ion source gas 1 (GS1), Ion source gas 2 (GS2), Temperature (Temp).

2.4. Statistical analysis

All graphs and statistical analyses were performed by using GraphPad Prism 9 (GraphPad Software; La Jolla, CA). Statistical analyses were carried out by using two-way ANOVA with Sidak's method (multiple comparisons) and t-test (two groups). Results are expressed as means ± S.E.M. *p ≤0.05, **p ≤0.01, ***p ≤0.001 corresponds significantly increased values compared to IV; whereas #p ≤0.05, ##p ≤0.01, ###p ≤0.001 corresponds significantly decreased values compared to IV.

3. Results

3.1. Biodistribution of 25 mg/kg DRV administered via IV and IN

The main purpose of this study is to compare the drug administration routes i.e., IV vs IN for in vivo biodistribution analysis of DRV as a free drug. First, we analyzed the distribution of a relatively high concentration of DRV (25 mg/kg) via IV and IN routes in the brain, liver, lungs, and plasma (Fig. 1, Supplementary Table 1a). Compared to IV delivery, the IN delivery of DRV significantly increased DRV accumulation in the brain, up to ∼ 2-fold (p ≤ 0.001; Fig. 1A), at 1h time point. However, it did not significantly facilitate DRV accumulation in the brain at later time points (3–12h). As expected, the accumulation of DRV in plasma was significantly higher via IV than IN route at 3h and 6h time points (p ≤ 0.001 or 0.05; Fig. 1G). Overall, compared to IV, the IN delivery of DRV significantly increased DRV accumulation cumulatively for all time points in the brain (19.74 ± 1.80 ng/g vs 12.84 ± 2.88 ng/g; p ≤ 0.05; Fig. 1B), while the IN delivery cumulatively decreased the DRV accumulation in plasma (320.56 ± 40.04 ng/g vs 899.12 ± 136.93 ng/g; p ≤ 0.001; Fig. 1H). On the other hand, we did not find a significant difference in the level of DRV accumulation in the liver or lungs with IN vs. IV administered DRV at any time point or cumulatively for all time points (Fig. 1C-D). Since inhalation procedure can lead to drug deposition in the lungs, we wanted to study the deposition of DRV in the lungs and comparison between the IV and IN routes. Our results showed similar accumulation of DRV in lungs via IV than IN routes at any time points or all time points cumulatively (Fig. 1E-F), suggesting that IN delivery would not cause lung toxicity.

Fig. 1.

Biodistribution of 25 mg/kg Darunavir (DRV) administered via intravenous (IV) and intranasal (IN) in Balb/c mice. DRV concentrations were analyzed in the brain (A, B), liver (C, D), and lungs (E, F), and plasma (G, H) at different time points at 25 mg/kg using IV and IN administrations. Concentrations of DRV were measured as ng/g in tissues and ng/ml in plasma (G, H). Statistical analyses were carried out by using ANOVA (multiple comparisons) and t-test (two groups). Results are expressed as means ± S.E.M. *p <0.05, **p <0.01, ***p <0.001 referred as significantly increased values, whereas #p <0.05, ##p <0.01, ###p <0.001 referred as significantly decreased values.

3.2. Biodistribution of 2.5 mg/kg DRV administered via IV and IN

In addition to high DRV concentration, we also administered low DRV concentration (2.5 mg/kg) via IV and IN routes to mice and analyzed the DRV concentration in brain, liver, lungs, and plasma at 1, 3, 6 and 12h time points (Fig. 2, Supplementary Table 1b). Compared to IV delivery of DRV, IN delivery of DRV increased DRV accumulation in the brain at early time points by ∼3-fold increase (p ≤ 0.05) at 1h and ∼2-fold increase (p ≤ 0.05) at 3h (Fig. 2A). Cumulatively, DRV accumulation also appeared to be increased with IN, however, no statistical significance was observed (12.08 ± 2.91 ng/g vs 6.91 ± 1.69 ng/g; Fig. 2B). As expected, IV administration of DRV showed significantly higher uptake of DRV both at cumulative and individual time points in the liver and plasma as compared to IN administration of DRV (plasma: 19.91 ± 12.65 ng/g vs 81.37 ± 22.04 ng/g, liver: 29.27 ± 4.17 ng/g vs 118.39 ± 28.13 ng/g; p ≤ 0.05, 0.01, or 0.001; Fig. 2C-D and 2G-H). The IN delivery of 2.5 mg/kg did not significantly increase lung DRV concentration compared to IV at individual or cumulative time points (Fig. 2E-F).

Fig. 2.

Biodistribution of 2.5 mg/kg Darunavir (DRV) administered via intravenous (IV) and intranasal (IN) in Balb/c mice. DRV concentrations were analyzed in the brain (A, B), liver (C, D), lungs (E, F), and plasma (G, H) at different time points at a dose of 2.5 mg/kg using IV and IN administrations. Concentrations of DRV were measured as ng/g in tissues and ng/ml in plasma. Results are expressed as means ± S.E.M. *p <0.05, **p <0.01, ***p <0.001 referred as significantly increased values, whereas #p <0.05, ##p <0.01, ###p <0.001 referred as significantly decreased values. The drug concentrations in this experiment were closer to the lower limit of quantification.

4. Discussion

This is the first time- and dose-dependent study that demonstrated enhanced delivery of DRV in the mice brain via IN compared to IV route. Furthermore, a significantly lower DRV accumulation in plasma and liver via IN than IV administration suggests a significantly reduced off-target effects and unwanted systemic side effects in the peripheral tissues/organs. Importantly, no further increased accumulation of DRV in the lungs via IN route, compared to IV administration, suggests that IN administration could also be safe for lungs.

In addition to high dose, delivery of low DRV dose via IN route also significantly improves brain DRV concentrations, with significantly lower systemic circulation and accumulation of DRV in liver or lungs. Moreover, a relatively higher DRV distribution in the brain at 2.5 mg/kg dose (∼5 ng/g tissue at 1h and ∼12 ng/g tissue cumulatively) than at 25 mg/kg dose (∼12 ng/g tissue at 1h and ∼20 ng/g tissue cumulative) is an important finding. The result suggests the benefit of the use of relatively low drug dose via IN administration in reducing neurotoxicity and off-target effects.

A quick delivery of DRV in brain using IN route likely results from direct absorption of DRV via olfactory tissues and bypassing the BBB. The IN delivery can directly absorb ∼50% the drug (depending upon the drug characteristics and concentration used) to the brain, and the remaining half could go through systematic circulation; some of which may further be delivered to brain [33,34]. Thus, accumulation of DRV in the brain, to some extent, at later time (12h; Fig. 2A) via IN route could be associated with initial delivery of DRV in plasma followed by its permeability to the brain via BBB.

Oral/parenteral delivery of antiretrovirals demonstrate poor CNS penetration, thus harboring HIV infection even with the antiretroviral treatment. Delivering antiretrovirals to CNS through IN route can be promising approach to achieve the effective drug concentrations and to target HIV in the brain. However, despite the evidence of showing better potential of drug targeting to the brain using IN route, there is no FDA-approved antiretroviral drug available for IN delivery. Additionally, extended release profile can maintain a consistent concentration of cART drug after administration, which can help the accumulation of cART drug via IN administration with lower dosing frequency [35].

Prior studies explored the delivery routes of saquinavir mesylate using nanoemulsion with an objective of effective drug targeting to the brain. The IN delivery of nanoemulsion-based saquinavir mesylate in the rats resulted in higher drug concentration in the brain than the IV delivery of nanoemulsion [36,37]. Moreover, the study did not show significant adverse effect in the cilia toxicity study on nasal mucosa. In another study, efavirenz-loaded nano lipid carrier demonstrated the therapeutic levels of the drug in the brain following IN delivery [38]. The results from lipid encapsulated drug showed a 10-fold increase in % drug targeting efficiency and >4-fold increase in % drug targeting potential for efavirenz compared to free efavirenz [38]. Similarly, pharmacokinetic study on the investigational drug DB213 showed better potential to deliver drug to the brain via IN administration with significantly high tissue distribution in the brain to plasma ratio as compared to IV administration [39]. Cubosome, the cubic phase nanoparticles reported as effective drug delivery vehicle for brain targeting [[40], [41], [42], [43]]. Furthermore, optimized IN delivery of cubosome loaded with saquinavir mesylate showed 12-fold higher bioavailability of the drug when compared with oral aqueous suspension. Taken together, these studies show great potential to deliver antiretrovirals, especially drug nanoformulations, using IN routes to achieve brain targeting for therapeutic alternative of HIV.

There is an urgent need for nanoparticle-based drug formulations coupled with effective IN delivery routes to facilitate FDA-approved antiretrovirals that could effectively cross/bypass the BBB and enhance drug bioavailability in the brain to curb neuroHIV. In prior studies, we reported that poly (lactic-co-glycolic acid) (PLGA)-based elvitegravir (EVG), an integrase inhibitor, nanoformulation enhanced intracellular EVG concentration up to 2.5-fold as compared to EVG alone in chronically HIV-infected human primary macrophages and microglia across the in vitro BBB as well as in animal model [[11], [12]]. These PLGA-EVG nanoformulation also showed higher efficacy for viral suppression compared with EVG alone in HIV-infected primary macrophages and microglia across the BBB and in HIV mice model [44]. Clearly, PLGA-EVG showed its ability to cross the BBB resulting in effective suppression HIV in HIV encephalitis (HIVE) mouse model [23]. To further improve antiretroviral drug concentration in the brain and subsequent viral suppression, we are in the process of formulating IN delivery of DRV using PLGA or other nanoparticles. These formulations are likely to further enhance drug delivery to the brain.

Limitations and Future Plans: The major sanctuary of HIV in brain is macrophages and microglia, and to some extent astrocytes [45]. Microglia, macrophages, and astrocytes are scattered throughout the CNS in a large, non-overlapping manner [46,47]. It is imperative to remove these infected cells from the brain, which are hinder from cART to achieve a cure. In our study we measured the concentration of DRV in whole brain as an overall goal to determine the enhancement of DRV concentration in brain. The future studies would constitute targeted delivery of DRV nanoformulations using specific antibody for macrophages and microglia that target HIV-infected perivascular macrophages and microglia, respectively. This is important because IN delivery will disperse the drug in whole brain, not specifically target HIV-infected macrophages and microglia. Thus, the combination of targeted nanoformulation and IN delivery would significantly enhance the prospects of utilizing antiretroviral drugs in effectively treating neuroHIV in the brain reservoirs macrophages and microglia. Furthermore, in this study we could not determine drug bioavailability, because studies with 4-time points yielded cumulative drug concentrations, not area under the curve (AUC). In the follow up study we will determine AUC using multiple time points (10 min–72 h) followed by drug bioavailability (AUC/dose) of DRV in IN nanoformulation. However, it's understood that bioavailability is direct proportion of cumulative brain concentrations.

Author contributions

S.K. conceived the project idea. A.K., L.Z., S.G. and N.S. performed the experiments. A.K., L.Z., and S.G. prepared the figures and analyzed the data. D.M. helped in the optimization of LC-MS/MS methods. K.P. helped in the analysis of results and sample preparation of LC-MS/MS. A.K., S.K., and L.Z. wrote the first draft of the manuscript. All authors reviewed the manuscript.

Funding

This study is supported by funding from the NIH grants DA047178 and DA125670 (SK).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article: