Kinetics of HTT lowering in brain of YAC 128 mice following single and repetitive intranasal dosing of siRNA packaged in chitosan-based nanoparticle

Abstract

This report describes the kinetics of Huntington’s Disease (HD) gene (HTT) lowering in brains of YAC 128 mice. Lowering (or “knock-down”) of HTT mRNA expression was achieved by intranasal administration of specially designed siRNA loaded into chitosan nanoparticles. Kinetic patterns of HTT lowering observed in different brain regions allowed calculation of cumulative lowering effects that result from multiple consecutive administrations. Mathematical modeling generated dosing schedules for approaching a steady knock-down effect and for prediction of magnitude and duration of HTT lowering. Kinetic modeling of HTT lowering with our algorithm will be useful in determining intranasal dosing schedules to produce chronic, therapeutically significant lowering effect of gene expression.

Graphical Abstract

1. Introduction

Small interfering RNA (siRNA) is a class of double-stranded RNA, typically 20–27 base pairs in length, designed to transiently lower expression of a target gene. Both viral and non-viral vectors have been used to deliver siRNA into cells, but the application of viral vectors in vivo has been reported to trigger unpredictable immune responses with occasional fatal consequences [1]. A cationic liposome such as Lipofectamine [2] can effectively transfect cells in vitro with high efficiency. However, in vivo application is not useful because of potential high toxicity and instability of cationic liposomes.

Recent studies from our laboratory have shown that chitosan nanoparticles loaded with siRNA can effectively and safely deliver the payload into mouse brain following intranasal instillation [3 – 6]. However, the total amount of siRNA administered with a single intranasal instillation may be insufficient to achieve the desired degree and duration of gene lowering that effectively prevents or slows disease progression. As the matter of fact, the gene lowering (also known as “knock-down”) is transient and requires repetitive dosing to maintain steady knock-down of the gene.

There are limited numbers of publications focused on kinetics of RNA interference associated with gene lowering. Some mathematical models were offered on kinetics of siRNA-mediated gene lowering [7 – 10]. Bartlett [11] applied mathematical modeling to investigate the kinetics of gene lowering, starting with delivery of siRNA and ending with extent of gene knock-down. This approach facilitated the design of siRNA-based strategies for both in vitro and in vivo treatment. Their proposed mathematical model did not consider various routes of administration, especially the intranasal route of administration, which has lately received increasing attention as a non-invasive approach for delivery of gene therapy [3, 12].

The primary advantage of intranasal administration is its simplicity and safety, allowing repetitive administrations for prolonged periods of time. This is especially important for application of nanocarriers that have a limited loading capacity for siRNA. By repetitive administration, the duration and magnitude of the therapeutic effect can be adjusted. In this report, we investigated the relationship between magnitude and duration of gene lowering in various brain regions as a function of dosing intervals. We administered siRNA packaged in nanoparticles and studied the effects of repeated intranasal administration on lowering duration and magnitude (degree of lowering of HTT expression) in a transgenic HD mouse model (YAC 128 mice bearing the human HTT gene) [13, 14]. We focused on striatum as the brain region where accumulation of mutant HTT protein causes significant neuronal dysfunction and pathology associated with the first motor manifestations of HD. The objective was to develop a mathematical model describing kinetics of gene lowering that can be used to calculate schedules of siRNA dosing that result in persistently reduced HTT expression. Our approach to kinetic modeling of HTT lowering is similar to the mathematical modeling used to determine key pharmacokinetic parameters of drug administration. Rather than measuring the time-course of drug concentrations in blood and target tissue, our approach is based on measuring the magnitude and duration of lowering of gene expression in specific brain tissues following single and repetitive intranasal administration of the nanoparticles.

2. Materials and Methods

2.1. Materials

Low molecular weight chitosan (mol. wt. 50,000–190,000 Da with 85% deacetylation) was obtained from Sigma-Aldrich (MO, USA). Mangafodipir Trisodium was purchased from U.S. Pharmacopeial Convention (MD, USA). The siRNA against HTT (Cy3-HTT-10150-P2VP) and non-target control siRNA (Cy3-NTC-P2VP-Chol) were synthesized in RNA Institute at the University of Massachusetts (MA, USA). This siRNA contained oligonucleotides to target exon 67 of HTT. The oligonucleotides were fully chemically modified for maximal stability and a passenger strand was labeled with Cy3 dye at 5’ end [15].

TaqMan™ Fast Advanced Master Mix and primers for qPCR were obtained from ThermoFisher (MA, USA), including HTT (assay ID: Hs00918174_m1) and PPIB as a reference gene (assay ID: Mm00478295_m1). All other chemicals and reagents used also were of analytical grade. Ultrapure distilled DNase/ RNase free water purchased from ThermoFisher (MA, USA) was used for all experiments.

2.2. Nanoparticle preparation

The siRNA was packaged into nanoparticles as described previously [3]. Fabrication of nanoparticles consisted of two steps including polyelectrolyte complexation and enrichment of the preparation. For polyelectrolyte complexation the chitosan was taken in concentration of 60 μM. The mixture of siRNA and Mangafodipir trisodium contained μM and 1 mM of each ingredient, respectively. The enrichment procedure carried out with Eppendorf Vacufuge centrifugal evaporator (Eppendorf, N.Y, USA) at 40 °C. The hydrodynamic size distribution and zeta potential in preparations were measured at 25 °C with a Malvern Zetasizer Nano ZS90 (MA, USA). Nanoparticle size in final enriched preparation was 123±2 nm with positive zeta potential of 47±3 mV. Enriched nanoparticle preparation contained 242 nmol/mL of siRNA.

2.3. Animals and treatment

All procedures with animals were conducted in compliance with National Institutes of Health guide for care and use of laboratory animals. All procedures with animals were performed in accordance with the Institutional Animal Care and Use Committee approved protocol (R IS00005853). Transgenic FVB-Tg (YAC128)33Hay/J and wild-type FVB/NJ mice were purchased from Jackson Laboratories (ME, USA) and breeding colonies were established in USF Comparative Medicine animal facility (protocol IS 00005132). Male Tg (YAC128) mice were bred with female FVB/NJ mice (background strain). The mice were housed under standard conditions with free access to water and food. For gene lowering experiments, we used early symptomatic 4-month-old female Tg (YAC128) mice and their wild-type (WT) littermates (FVB/NJ). Total number of animals used in gene lowering experiments including non-target control was 50.

Mice were treated with siRNA loaded nanoparticles via intranasal administration. Non-target siRNA was used for control experiments. Before intranasal administration mice were lightly anesthetized with isoflurane. For intranasal administration we applied standard method of instillation accepted elsewhere [16, 17]. Each mouse was gently grasped by the back of the neck with abdomen facing upwards while 6 μl of nanoparticles volume was instilled in each nostril dropwise over 30 sec with 2 min break between each pair instillations. Because duration of administration session was 4 min and lowering effect lasts several hours, such administration session was considered as single administration. In total, each animal received 24 μl of nanoparticles during 4 min of single administration session. The siRNA dose was 5.8 nmol correspondently to the dose used in previous study [3].

2.4. Tissue collection

Tissue sampling from olfactory bulb (OB), cerebral cortex (CX), corpus striatum (ST) and hippocampus (HP) was performed by micro-punching of coronal sections of brain described earlier [3]. Brain coronal sections of 1 mm thickness were prepared at minus 10 °C in a cryostat chamber using a Brain Matrices mouse brain mold (Braintree Sci., Inc., MA, USA). The coronal sections were mounted on Superfrost/Plus microscope slides (Fisherbrand, PA, USA) and immersed in cold RNAlater stabilization solution (ThermoFisher, MA, USA) for 30 min, then stored at minus 70 °C until dissection of specific brain loci. The harvesting of brain tissue from specific loci of each coronal section was performed under a stereo microscope (WILD Heerbrugg, Switzerland) with Brain Tissue Punch Set (Vibratome, MO, USA). Thermo-electric Cold Plate (Oven Ind., Inc., PA, USA) was used as a microscope stage. The tissue mounted on slides sections were dissected at minus 15 °C permitting precise punching without tissue rupture. Collected samples were stored in the 1.5 mL plastic tubes at minus 80 °C until further processing. Dissection was controlled under DEI-470 (Optronics Eng., CA, USA) video camera connected to computer. Selection and identification microdissection sites performed with guidance of the mouse-brain atlas [18].

2.5. Methods for evaluation of HTT lowering (“knock-down”)

Levels of mRNA expression were measured with QuantStudio 3 (ThermoFisher, MA, USA). TaqMan™ Fast Advanced Master Mix (ThermoFisher) was used for quantitative Real-Time PCR (qPCR) carried out in accordance to manufacturer protocol. Primers for qPCR were obtained from ThermoFisher, including HTT (assay ID: Hs00918174_m1) and PPIB (assay ID: Mm00478295_m1). Expression levels for huntingtin mRNA were normalized to PPIB (peptidylprolyl isomerase B) endogenous control. Total RNA was extracted with RNAeasy Mini Kit (Qiagen, CA, USA). The RNA was reverse transcribed using Invitrogen™ SuperScript™ III Reverse Transcriptase kit and Invitrogen™ Oligo(dT) 20 Primer (Fisher Scientific, PA, USA).

The ddCt method [19] was used to calculate the percent of remaining HTT mRNA and the percent of lowering effect was expressed as 100% minus remaining % of HTT mRNA. HTT gene lowering in mouse brain resulted from multiple intranasal administrations of nanoparticles carrying anti-HTT siRNA expressed by following mathematical function:

$$KD(t)=M{\displaystyle \sum }_{i=1}^n{W}^{-k_i}$$

Where KD(t) is cumulative lowering effect observed at time point t; M represents magnitude of lowering effect; W determines shape of curve, and n is number of consecutive administrations. The exponent factor k_i is calculated as k_i = (c + a_i − t)², where c = culminating time; a_i = administration time for each of consecutive administration (for example, first administration time a₁ is 0 hr, second administration time a₂= 24 hr)

2.6. Statistics

HTT lowering effects of siRNA were calculated by averaging rt-PCR tests for group of 5 animals (n=5) and expressed as Mean ± SEM. Means and SEM, were calculated with GraphPad Prism v. 8.1.0 (GraphPad Software, CA, USA, and Microsoft Excel for Office 365 (MSO 16.0.12527.21378).

3. Results and Discussion

To evaluate effects of HTT mRNA lowering (or” knock-down”) across the brain, we dissected brain tissue into four regions including olfactory bulb (OB), hippocampus (HP), striatum (ST) and cortex (CX). The HTT lowering data in different brain regions at 6, 12, 24, 48 and 72 hrs post administration is presented in Table 1. The magnitude and duration of HTT lowering varied across the four brain regions. Changes in the magnitude of HTT lowering presented in Table 1 correlated strongly with values calculated with our mathematical algorithm (Table 2). This algorithm generated kinetic lowering patterns useful for accurate calculation of magnitude and duration of gene expression. Such calculations would be impossible when using a small number of experimental time points and/or limited number of animals for study.

Table 1.

Kinetics of HTT lowering in different regions of mouse brain expressed as lowering percent (KD±SEM) averaged for groups of 5 animals. SEM represents standard error of mean.

Post administration time (hrs)	Average lowering effect by regions (mean ± SEM)*
	OB	HP	ST	CX
	6	3.66±2.01	0.99±4.74	−1.59±3.16	−1.53±1.12
12	2.23±4.26	0.49±4.33	7.41±0.88	−0.09±3.50
24	20.80±5.50	2.17±4.18	12.64±2.39	0.82±2.70
48	17.77±3.07	18.49±2.25	20.80±4.76	11.03±3.70
72	5.12±3.56	12.88±6.36	7.93±6.04	12.75±3.33

^*Lowering effect was calculated from qPCR threshold cycle number (Ct) by following formula: KD = (1 – 2^-ddCt)×100% where ddCt = (Ct _htt – Ct _ppib)_treatment – (Ct _htt – Ct _ppib)_control

Table 2.

Fitting of experimental data presented in Table 1 with KD(t) curves generated with proposed formula.

Fitting variables:	OB	HP	ST	CX
Curve shape factor, W	1.0038	1.0021	1.0014	1.0023
Culminating time, C	35.3	56.46	44.4	61.34
Magnitude of the effect, M	33.21	21.5	21.64	16.48
Calculated KD(t):
@ 6 hr	1.26	0.09	2.77	0.02
@12 hr	4.21	0.33	5.00	0.07
@24 hr	20.43	2.31	12.11	0.72
@48 hr	17.97	18.48	21.26	11.04
@72 hr	0.19	12.89	7.48	12.75
Correlation coefficient, R2	0.9647	0.999	0.9538	0.9958

The kinetics of the HTT lowering effect resulting from a single intranasal dose of siRNA-loaded nanoparticles is shown in Fig. 1. There were four distinct patterns of HTT lowering (also shown in Table 2). The kinetic pattern of HTT mRNA lowering traced a bell-shaped curve with different culminating times and magnitudes of effect (Fig. 1). A maximal HTT lowering of 30% was attained at 36 hrs in the OB. HTT lowering in other regions of brain developed at a slower pace as compared to OB. In ST the maximal knock-down effect was delayed by 10 hr compared to OB. Kinetics of HTT lowering in HP and CX was even slower indicating delays of 18 and 34 hr, respectively. The magnitude of the lowering effect was also highest in OB. In ST it reached 80% of the OB magnitude, and 60% and 47% of OB magnitude in HP and CX, respectively. Assuming that velocity of degradation of HTT mRNA in brain regions is approximately equal, the time course of HTT lowering demonstrates the presence of four distinct patterns that may reflect differential distribution and accumulation of the nanoparticles across different brain regions. Mathematical modeling allowed prediction of the cumulative effect on gene lowering attained following two consecutive intranasal administrations. For illustration, ST was selected because it is the brain region that plays a crucial role in early motor manifestations of HD. As can be seen from Fig. 2 and 3, two consecutive doses administered 6 hrs apart (Fig. 2) doubled the magnitude of lowering effect in comparison with single administration. Duration of the knock-down effect was also extended from 40 hrs to 56 hrs (Fig 2). Area under curve of the cumulative effect is equal to sum of two single administrations.

Fig. 1.

Kinetics of HTT lowering (KD,%) in four regions of YAC 128 transgenic mouse brain following a single session of intranasal administration of nanoparticles bearing anti-HTT siRNA. KD, % was calculated by hourly intervals using the variables W, C and M obtained for each brain region from fitting of experimental data as presented in Table 2.

Fig. 2.

Kinetics of lowering effect of HTT in ST of YAC 128 mice resulting from two consecutive intranasal administration 6 hrs apart of nanoparticles carrying siRNA. The resulting cumulative effect depicted by grey line. Blue line represents kinetics predicted for the first administration (a₁=0) and orange curve describes kinetic for second administration (a₂=6 hrs). Calculations were performed assuming the following parameters: W=1.0014; C=44.4 and M=1.

Fig. 3.

Kinetics of lowering effect of HTT in ST of YAC 128 mice following two consecutive intranasal administration 24 hrs apart of nanoparticles carrying siRNA. The resulting cumulative effect is depicted by grey line. Blue line represents kinetic predicted for the first administration (a₁=0) and orange curve describes kinetic for second administration (a₂=24 hrs). Calculations were performed assuming the following parameters: W=1.0014; C=44.4 and M=1.

Two repetitive administrations 24 hrs apart (Fig. 3) increased the magnitude of cumulative effect in HTT lowering 1.5-fold as compared to single administration of nanoparticles carrying siRNA. However, the magnitude of effect was less compared to administration of two doses 6 hrs apart (Fig. 2). With this schedule of administration, duration of cumulative lowering effect (measured at 50% level of lowering) was extended to 65 hrs in comparison to 40 hrs caused by a single dose.

Thus, the duration and magnitude of cumulative effect is dependent on time between administrations (Fig. 4). The magnitude of cumulative effect is highest when there is minimal time between administrations and returns to the level of single dose administration when the time between administrations is increased beyond 45 hrs. However, duration of cumulative effect increases as time between administrations is extended.

Fig. 4.

The cumulative effect of HTT lowering in ST of YAC 128 mice plotted against administration time interval. The magnitude of lowering (blue line) and duration of lowering (orange line) produced by two consecutive doses administered at different intervals.

In our previous report [3] we used 4 consecutive administrations with the following time intervals: a₁ = 0 hrs; a₂ = 6 hrs; a3 = 24 hrs; a₄ = 30 hrs. Brain tissue was harvested 48 hr after administration of the first dose. Modeling the anticipated magnitude of the lowering effect showed a 2.63-fold increase at the time of euthanasia that is consistent with experimental data [3], where lowering effect in ST reached 62±5%. These results confirm the validity of our proposed mathematical model.

A graphical method for modeling the time course of HTT suppression in adult female rhesus monkeys following direct infusion of siRNA into brain parenchyma has been reported but has limitations [20] This graphical method requires selection of a single curve from a series of alternative curves generated by fitting experimental data. This approach is not appropriate for prediction of HTT suppression beyond the experimental time frame and, therefore cannot be used to determine optimal dosing schedules to achieve the desired steady-state level of HTT knockdown.

The mathematical algorithm modeling the kinetics of gene lowering described here is useful for generating dosing schedules to achieve a steady level of HTT lowering as illustrated by Fig. 5. Unlike classical pharmacokinetic studies, which measure the time course of drug concentrations over time in blood and target tissue, here we relied on measurement of a specific biological effect (HTT mRNA lowering) and how it changes over time in various brain regions. The parameters of gene lowering depend on dynamic cellular physiological processes such as a) rate and mechanism of transport into brain from nasal mucosa, b) time course of distribution across brain regions, c) rate of release of the payload (siRNA) in cells, d) rate of clearance or metabolism of siRNA, and e) rate of resynthesis of HTT protein. To achieve a consistent and steady lowering of gene expression, the frequency of intranasal administration of the NPs loaded with siRNA depends on the balance between extent of gene lowering and rate of new HTT expression.

Fig. 5.

Example of dosing schedule intended to maintain steady lowering effect in ST with multiple administrations. Depicted here are 4 consecutive administrations made 24 hrs apart (blue, orange, grey and yellow lines) with green line representing cumulative effect. Number of administrations can be extended to prolong duration of HTT lowering. Calculations were performed assuming the following parameters: W=1.0014; C=44.4 and M=1.

The distinct kinetic curves of HTT lowering observed in specific brain regions may be explained, in part, by the different mechanisms of transport and distribution of nanoparticles to brain regions following instillation into the nasal cavity. Briefly, nasal epithelium is innervated by olfactory nerves (ON) and branches from the Vth cranial nerve, the trigeminal nerve (TN). Following intranasal instillation, NPs can be transported directly into brain, bypassing the blood-brain-barrier by two mechanisms: 1) transcellular (uptake into olfactory nerve terminals with transport to cell bodies of the olfactory bulb), and 2) passage into the perineural space created by the olfactory nerve ensheathing cells and the peri-vascular space, ultimately reaching the cerebrospinal fluid [21].

We speculate that transcellular transport by ON of our nanoparticles occurs faster, accounting for the relatively rapid onset of gene lowering in the olfactory bulb and striatum. The neural route starts with receptor-mediated uptake (divalent metal transport) of manganese-containing nanoparticles by ON and TN terminals in the nasal epithelium. ONs project to olfactory bulb neurons, which in turn project to ventral striatum and ventral pallidum via the olfactory tubercle [22]. The olfactory tubercle pathways also project to the piriform cortex, also known as olfactory cortex, located in ventral forebrain. Of note, we did not measure the kinetics of gene lowering in brain stem or cerebellum, but we postulate that transport of NPs to pontine nuclei via the trigeminal nerve will result in relatively rapid onset of lowering in that region.

We also speculate the relatively slower onset of lowering may be a function of nanoparticle transport into the CSF via perineural and perivascular space. Within the CSF, nanoparticles are distributed by bulk CSF flow and infiltration into the extracellular milieu where they are taken up by neurons and glial cells in cerebral cortex and hippocampus. Finally, the magnitude and duration of lowering in each brain region will also be impacted by cellular processes such as the rates of siRNA release from NPs into the target cells, clearance of the siRNA and the baseline rate of HTT mRNA expression in the region.

In conclusion, we modeled the kinetics of gene lowering to determine a dosing schedule that will produce long-term therapeutically significant gene knock-down. An optimal dosing schedule to achieve a steady level of gene knock-down will be essential for translation of intranasal delivery of gene therapy to human clinical trials.

Highlights.

• The time-course and magnitude of huntingtin gene (HTT) silencing was determined in the transgenic YAC128 mouse model of HD following intranasal administrations of siRNA loaded into chitosan nanoparticles.

• A mathematical algorithm for generating kinetic patterns of HTT silencing in different brain regions allowed calculation of a cumulative silencing effect achieved by multiple consecutive administrations.

• Mathematical modeling is useful for selecting a schedule of intranasal gene therapy that results in a steady knock-down effect and for prediction of magnitude and duration of the gene silencing.