Advances in the development of mRNA (LNP) vaccines have made it possible to obtain two FDA approved vaccines (Pfizer/ BioNTech and Moderna) against the SARS-CoV-2 virus in less than a year [1, 2]. The LNP–mRNA-based medications can be used for both treatment of a number of socially significant disorders and as vaccines for prevention of infections caused by many pathogens. The mRNA–LNP platform flexibility is due to the possibility of specific selection of the antigenic sequence comprised in the mRNA molecule, it is also due to different variants of the lipid composition and their ratios in LNPs that can modulate the mRNA vaccine efficiency and immunogenicity [3]. The Pfizer/BioNTech and Moderna lipid particles comprise charged ionized lipids, neutral ionized lipids, poly(ethylene glycol)- containing lipids, cholesterol, and distearoylphosphatidylcholine (DSPC) [4]. LNPs ensure mRNA–LNP internalization into the cell and play an adjuvant role, stimulating a moderate increase in the injection site inflammation. In has been shown that different variants of ionizable lipids recognized by the toll-like receptor 4 (TLR4) play a central role in the induction of inflammation caused by LNPs [5]. Furthermore, the mRNA molecule being a vaccine component can exert pro-inflammatory activity via TLR-3,7,8, RIG-I, MDA5 [6, 7]. Moderate pro-inflammatory activity contributes to effective antigen presentation of the antigen-presenting cells, as well as to the humoral and T-cell immunity formation. However, inflammation may sometimes cause adverse effects. In particular, recent studies have shown that LNPs cause severe injection site inflammation, have a broad biodistribution profile, and are found in multiple tissues of the body, including the brain [4, 8]. Uninhibited crossing the blood-brain barrier together with pro-inflammatory activity can cause adverse effects in the form of immune activation in the central nervous system. The study was aimed to perform the dynamic assessment of neuroinflammatory markers in the prefrontal cortex and hypothalamus of the Balb/c mice after administration of various mRNA–LNP vaccine doses.

METHODS

Experimental design

The conventional experiment involved 75 adult Balb/c males (age 9–10 weeks, body weight 19–22 g) obtained from the Rappolovo Breeding Nursery of the Russian Academy of Medical Sciences (St. Petersburg, Russia) and kept at the Center of Experimental Pharmacology, St. Petersburg State Chemical and Pharmaceutical University, under fixed lightning conditions (12.00 : 12.00 h). The animals had free access to the standard food (granules) and water. The animals were distributed into the study groups by randomization before the study. Intramuscular injections of 30 μL of various doses of mRNA–LNP (three concentrations: 5, 10, and 20 μg of RNA) or control (empty) LNPs in phosphate buffer were performed. The animals inhaled the 2.0% isoflurane (Laboratories Karizoo, S.A.; Spain) mixed with oxygen for 5 min and were subsequently decapitated within 4, 8, and 30 h after administration of the particle suspension (fig. 1). The samples of the hypothalamis and prefrontal cortex (PFC) were obtained as earlier reported [9]. The same volume (30 μL) of phosphate buffer was administered to the control animals. Five animals per experimental point were used in each group.

Cloning

Amplification of the target gene comprising the 5'-UTR Moderna (gggaaataagagagaaaagaagagtaagaagaaatat aagaccccggcgccgccacc) encoding the firefly (Photinus pyralis) luciferase and the 3'-UTR Moderna (gctggagcctcggtggcctagcttcttgccccttgggcctccccccagcccctcctccccttcctgc acccgtacccccgtgtctttgaataaagtctgagtgggcggca) sequences was performed via linking together three fragments by the overlapping primer-based PCR. Then the resulting fragment was incubated with the EcoRI and BglII restriction endonucleases, purified from agarose gel and ligated to the pSmart commercial vector (Lucigen; USA) prepared by the same method. The vector comprised a polyA-tail with the size of 110. A NEB-stable E. coli strain (New England Biolabs; UK) was used for transformation. Clones were selected from the colonies by PCR, and the sequence of the insert was confirmed by sequencing. To generate the verified plasmid, E.coli was grown in the incubator shaker at 30 °C and 180 rpm. Then plasmid DNA was extracted from bacterial cells using the QIAGEN Plasmid Maxi Kit (Qiagen; USA). The resulting plasmid preparation was linearized by the unique SpeI restriction site and subsequently visualized in agarose gel.

In vitro mRNA transcription

In vitro transcription was carried out in the buffer solution containing 20 mmol of DTT, 2 mmol of spermidine, 80 mmol of HEPES-KOH (pH 7.4), 24 mmol of MgCl₂. The reaction mixture also contained 3 mmol of each ribonucleoside triphosphate (Biosan; Russia), 12 mmol of anti-reverse cap analog (ARCA) (Biolabmix; Russia). Other components per 100-μL reaction volume: 40 units of the RiboCare ribonuclease inhibitor (Evrogen; Russia), 500 units of the T7 RNA polymerase (Biolabmix; Russia), 5 μg of the linearized plazmid, and 1 μL of the enzyme mix from the RiboMAX Large Scale RNA Production System kit (Promega; USA) as the source of inorganic pyrophosphatase. The reaction was carried out for 2 h at a temperature of 37 °C, then another 3 mmol of each ribonucleoside triphosphate were added to the reaction and incubated for 2 h. DNA was hydrolyzed using the RQ1 nuclease (Promega; USA), RNA was precipitated by adding LiCl to a concentration of 0.32 mol and EDTA (pH 8.0) to a concentration of 20 mmol with subsequent incubation on ice for an hour. Then the solution was centrifuged for 15 min (25,000 g, 4 °C). RNA precipitate was washed with 70% ethanol, diluted in the ultrapure water and once more precipitated by alcohol using the standard method. RNA concentration was defined by spectrophotometry based on absorbance at a wavelength of 260 nm.

Formulation of LNPs containing mRNA

Encapulation of mRNA into lipid nanoparticles was performed by mixing the 0.2 mg/mL mRNA aqueous solution (10 mmol citrate buffer, pH 3.0) with the alcohol solution of the lipid mixture in the microfluidic cartridge using the NanoAssemblr Benchtop system (Precision Nanosystems; USA). The lipid mixture contained the following components: ALC-0315 ionizable lipidoid (BroadPharm; USA), distearoylphosphatidylcholine (DSPC) (Avanti Polar Lipids; USA), cholesterol (Sigma-Aldrich; USA), DMG-PEG-2000 (BroadPharm; USA) in a molar ratio (%) of 46.3 : 9.4 : 42.7 : 1.6. The amount of lipids per unit of mRNA was calculated based on the following ratio: N/P = 6 (ALC-0315 ionizable lipidoid/mRNA base). To generate particles of the desired size, the aqueous and alcohol phases were mixed in a ratio of 3 : 1 v/v with the total mixing speed of 10 mL/min. After mixing the phases the resulting water-alcohol particle suspension was dialyzed in 300 volumes of phosphate buffered saline (pH 7.4, 18 h, +15 °C). After dyalisis the particle suspension was concentrated using the Amicon Ultra-4 10,000 Da molecular weight cutoff filter. Then particles were filtered through the filter with the 0.22 μm PES membrane (Merck; USA) and stored at 4 °С. Empty LNPs were obtained by mixing the 10 mmol citrate buffer (pH 3.0) with the lipid mixture alcohol solution in the microfluidic cartridge by the same method that was used to obtain the mRNA-loaded LNPs.

After filtration, the quality of the particles generated was assessed based on two parameters: mRNA load and particle size. The concentration of mRNA loaded into lipid nanoparticles was defined based on the differences in the fluorescence signal levels obtained for the particle suspension stained with the RiboGreen reagent (Thermo Fischer Scientific; USA) before and after the particle disruption. The Triton X-100 detergent (Sigma-Aldrich; USA) was used to disrupt the particles. The LNP size was defined by the dynamic light scattering method in the Zetasizer Nano ZSP system (Malvern Panalitycal; USA).

Estimation of gene expression in the brain

Total RNA was extracted from the PFC and hypothalamus using the kit for column-based RNA isolation (Biolabmix; Russia) in accordance with the manufacturer's protocol. RNA concentration and purity were assessed with the NanoDrop OneC spectrophotometer (Thermo Scientific; USA).
To carry out the reverse transcription reaction, 500 ng of RNA and the ОТ-M-MuLV-RH reverse transcription kit (Biolabmix; Russia) with random hexanucleotide primers were used. The resulting cDNA was used to assess gene expression. Expression levels of the genes encoding pro-inflammatory cytokines and interleukins (Il1β, Tnfα), marker genes of microglia (Aif1) and astroglia (Gfap) activation were assessed as neuroinflammation markers. The study involved the use of quantitative PCR with fluorescent Taq-man probes. The sequences of primers and probes are provided in table.
The expression was assessed relative to mRNA of the housekeeping gene (Gapdh). PCR was carried out using the BioMaster HS-qPCR (2×) kit (Biolabmix; Russia) in the Real- Time CFX96 Touch system (Bio-Rad Laboratories; USA) in accordance with the following protocol: 95 °С for 15 s, 60 °С for 20 s. Three iterations of all tests per cDNA sample were performed. The expression was quantified by the ΔΔСt method.

Statistical analysis

Statistical processing of the results was performed by ANOVA (the “group” and “time after administration” were used as factors) and Fisher’s least significant difference (LSD) test as a post hoc test. The differences between the experimental groups were considered singnificant at p < 0.05, while at the level of trends these were considered significant at p < 0.1. Data analysis was performed using the Statistica 8.0 software package (Statsoft Inc.; USA).

RESULTS

The findings show that various mRNA–LNP vaccine doses induce activation of Aif1 in the hypothalamus (fig. 2), but not in the prefrontal cortex (fig. 3). The two-way analysis of variance (ANOVA) made it possible to reveal significant effects of the “group” and “time after administration” factors on the Aif1 expression in the hypothalamus (F(4.70) = 2.866 at p = 0.032; F(2.72) = 4.246 at p = 0.019). In the groups of mice that received 10 μg of mRNA and 20 μg of RNA as part of the mRNA–LNP vaccine, the expression of Aif1 mRNA within 7 h after the vaccine administration was about 80% higher than in the control group that received phosphate buffer (p > 0.05). It is interesting to note that the groups that received 5 μg of RNA as part of the mRNA–LNP vaccine and empty LNPs (with no mRNA) also showed elevated espression of Aif1 (by 40–55%) within 8 h, however, these differences were non-significant. No differences in the hypothalamic Aif1 expression between animals of different groups were observed 30 h after the vaccine administration. No significant effects of the “group” factor or the interaction of the “group” and “time after administration” factors on the expression of other assessed genes in the hypothalamus (Tnfα, Il1β, Gfap) and gene expression in the prefrontal cortex were revealed. Thus, we observed moderate mRNA–LNP effects on the neuroinflammation associated with the elevated expression of the markers of active microglia in the hypothalamus, but not in the prefrontal cortex. Furthermore, these effects were dose-dependent.

Comparison of gene expression at various time points between animals of various groups after administration of the mRNA–LNP vaccine showed that Il1β expression was dramatically increased 4 h after vaccination in both hypothalamus and prefrontal cortex of certain animals in the groups that received 10 μg of mRNA and 20 μg of RNA as part of the mRNA–LNP vaccine. However, no such effects were observed in the later measurement points. Despite the profound effects on the Il1β, these differences were non-significant, since only a few animals in the groups showed a pronounced response. Such results demonstrate heterogeneity of the response to the mRNA–LNP vaccine associated with individual characteristics of the animals.

The effect of the “time after administration” factor on the Gfap and Tnfα expression in the hypothalamus was revealed (F(2.72) = 10.179 at p < 0.0001; F(2.72) = 5.181 at p = 0.008). The Gfap expression decreased within 8 h in all experimental groups, however, it increased in 30 h. It is interesting that the Tnfα expression also increased in 30 h after vaccination compared to the levels observed within 4 h in the majority of experimental groups. Such results suggest that mice in the experimental group develop the second wave of pro-inflammatory activation involving astrocytes and interleukin TNFα.

DISCUSSION

The findings show that mRNA–LNP vaccines with the mRNA doses of 10–20 μg are capable of increasing the Aif1 expression within 8 h in the hypothalamus, but not in the prefrontal cortex. We have found that experimental groups demonstrate the differences in the Gfap, Il1b, Tnfα expression levels measured at various time points in the hypothalamus, which is also an indirect evidence of the fact that the expression levels of these genes may be correlated to the mRNA–LNP vaccine administration.

The mRNA–LNP vaccine can cause both local and systemic inflammation [4, 8]. Inflammation can be caused by various vaccine components: mRNA molecules, lipids forming part of LNPs or protein product encoded by mRNA. The mRNA–LNPs most often transfect cells near the injection site, after that LNPs are rapidly transported to the proximal lymph nodes by passive drainage and are also actively transported by the professional antigen-presenting cells and neutrophils [10, 11]. Then mRNA–LNP can reach any cell of the body via systemic circulation; low amounts of mRNA–LNP are found in the brain, thus suggesting its capability of crossing the blood-brain barrier [12, 13].

It is known that peripheral inflammatory stimuli can also cause immune response in the brain that results in activation of astrocytes, the main immunocompetent cells of the brain [14]. Because of their cytokine-producing and phagocytic activity, these cells affect the development and maturation of the CNS structures [15], participate in the normal formation and development of neural circuits during onthogenesis [16], maintain the pool of neurons, mediate synapse maturation and reduction, thereby regulating the number of synapses and receptor expression [17].

Thus, the signs of microglia activation we have found in certain experimental groups may be both evidence of mRNA– LNP directly crossing the blood-brain barrier and triggering neuroinflammation, and the result of the increasing peripheral inflammation. Since our study does not involve assessment of the peripheral immune activation parameters, we cannot answer this question explicitly.

Significant differences in the Aif1 expression revealed 8 h after immunization are consistent with the data showing that the peak of microglia activation falls between 6–24 h after induction of inflammation [14, 18–20]. At the same time, the peak of cytokine activation after induction by inflammatory agents, such as bacterial lipopolysaccharide or the synthetic analog of double-stranded RNA (Poly I:C), falls between 1.5–3.0 after administration of inflammatory mimetics. That is why the lack of significant effects on the expression of Il1β and Tnfα observed across the groups may be due to the fact that peak activation of gene expression is passed. At the same time, a number of studies show that elevated cytokine levels in the brain and periphery may persist up to 24 h after inflammation induction by mimetics.

In our study we assessed the expression of pro-inflammatory genes in two brain structures. The more pronounced effects were observed in the hypothalamus, while prefrontal cortex showed no significant alterations. The hypothalamus is an important brain structure that functions as a metabolic center responsible for regulation of multiple fundamental physiological processes involved in metabolism of the whole body, including food intake, regulation of appetite, energy consumption; thus, the hypothalamus plays a crucial role in systemic homeostatic regulation [22]. Clinical data have shown that various stimuli, such as peripheral inflammation or the increased intake of saturated fatty acids, may cause neuroinfllammation in this brain structure [23–25]. Furthermore, the hypothalamus contains various cell populations of microglia and astroglia [26]. Taken together, these data show that the hypothalamus may be a kind of the peripheral inflammation sensor and respond to pro-inflammatory signals more actively than the prefrontal cortex.

CONCLUSIONS

The mRNA–LNP vaccine can activate the hypothalamic Aif1 expression 8 h after vaccination in a dose-dependent manner. However, no significant effects of mRNA–LNP vaccines on the gene expression have been found in the prefrontal cortex. Despite the fact that alterations in the Aif1 expression observed within 30 h after vaccination are non-significant, these findings show that mRNA–LNP vaccine can induce neuroinflammation. Further experiments involving larger groups of animals and focused on assessing the parameters of peripheral inflammation and broader analysis of neuroinflammation involving the use of immunoassays and immunohistochemistry for assessment of pro-inflammatory agents and microglial cell morphology in the hypothalamus and other brain structures are required to understand the mechanisms underlying the mRNA–LNP vaccine capability of inflammation stimulation.