Midazolam impacts acetyl—And butyrylcholinesterase genes: An epigenetic explanation for postoperative delirium?

Abstract

Midazolam is a widely used short-acting benzodiazepine. However, midazolam is also criticized for its deliriogenic potential. Since delirium is associated with a malfunction of the neurotransmitter acetylcholine, midazolam appears to interfere with its proper metabolism, which can be triggered by epigenetic modifications. Consequently, we tested the hypothesis that midazolam indeed changes the expression and activity of cholinergic genes by acetylcholinesterase assay and qPCR. Furthermore, we investigated the occurrence of changes in the epigenetic landscape by methylation specific PCR, ChiP-Assay and histone ELISA. In an in-vitro model containing SH-SY5Y neuroblastoma cells, U343 glioblastoma cells, and human peripheral blood mononuclear cells, we found that midazolam altered the activity of acetylcholinesterase /buturylcholinesterase (AChE / BChE). Interestingly, the increased expression of the buturylcholinesterase evoked by midazolam was accompanied by a reduced methylation of the BCHE gene and the di-methylation of histone 3 lysine 4 and came along with an increased expression of the lysine specific demethylase KDM1A. Last, inflammatory cytokines were not induced by midazolam. In conclusion, we found a promising mechanistic link between midazolam treatment and delirium, due to a significant disruption in cholinesterase homeostasis. In addition, midazolam seems to provoke profound changes in the epigenetic landscape. Therefore, our results can contribute to a better understanding of the hitherto poorly understood interactions and risk factors of midazolam on delirium.

Full text

RESEARCH ARTICLE
Midazolam impacts acetyl—And
butyrylcholinesterase genes: An epigenetic
explanation for postoperative delirium?
Katharina RumpID*, Caroline Holtkamp, Lars Bergmann, Hartmuth Nowak,
Matthias Unterberg, Jennifer Orlowski, Patrick Thon, Zainab Bazzi, Maha Bazzi,
Michael Adamzik, Bjo
¨rn Koos, Tim Rahmel
Department of Anesthesiology, Intensive Care Medicine and Pain Therapy, University Hospital
Knappschaftskrankenhaus Bochum, Ruhr-University Bochum, Bochum, Germany
*Katharina.k.rump@rub.de
Abstract
Midazolam is a widely used short-acting benzodiazepine. However, midazolam is also criti-
cized for its deliriogenic potential. Since delirium is associated with a malfunction of the neu-
rotransmitter acetylcholine, midazolam appears to interfere with its proper metabolism,
which can be triggered by epigenetic modifications. Consequently, we tested the hypothesis
that midazolam indeed changes the expression and activity of cholinergic genes by acetyl-
cholinesterase assay and qPCR. Furthermore, we investigated the occurrence of changes
in the epigenetic landscape by methylation specific PCR, ChiP-Assay and histone ELISA. In
an in-vitro model containing SH-SY5Y neuroblastoma cells, U343 glioblastoma cells, and
human peripheral blood mononuclear cells, we found that midazolam altered the activity of
acetylcholinesterase /buturylcholinesterase (AChE / BChE). Interestingly, the increased
expression of the buturylcholinesterase evoked by midazolam was accompanied by a
reduced methylation of the BCHE gene and the di-methylation of histone 3 lysine 4 and
came along with an increased expression of the lysine specific demethylase KDM1A. Last,
inflammatory cytokines were not induced by midazolam. In conclusion, we found a promis-
ing mechanistic link between midazolam treatment and delirium, due to a significant disrup-
tion in cholinesterase homeostasis. In addition, midazolam seems to provoke profound
changes in the epigenetic landscape. Therefore, our results can contribute to a better under-
standing of the hitherto poorly understood interactions and risk factors of midazolam on
delirium.
1. Introduction
Midazolam is the most abundantly used benzodiazepine in anesthesia and emergency medi-
cine [1]. Due to its amnestic and anxiolytic effects, midazolam is considered as a favorable
choice for premedication [2,3]. However, the use of benzodiazepines especially midazolam is
associated with postoperative complications such as cognitive impairment and delirium [4].
PLOS ONE
PLOS ONE | https://doi.org/10.1371/journal.pone.0271119 July 8, 2022 1 / 12
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPEN ACCESS
Citation: Rump K, Holtkamp C, Bergmann L,
Nowak H, Unterberg M, Orlowski J, et al. (2022)
Midazolam impacts acetyl—And
butyrylcholinesterase genes: An epigenetic
explanation for postoperative delirium? PLoS ONE
17(7): e0271119. https://doi.org/10.1371/journal.
pone.0271119
Editor: Israel Silman, Weizmann Institute of
Science, ISRAEL
Received: May 9, 2022
Accepted: June 23, 2022
Published: July 8, 2022
Peer Review History: PLOS recognizes the
benefits of transparency in the peer review
process; therefore, we enable the publication of
all of the content of peer review and author
responses alongside final, published articles. The
editorial history of this article is available here:
https://doi.org/10.1371/journal.pone.0271119
Copyright: ©2022 Rump et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited.
Data Availability Statement: All relevant data are
within the paper and its Supporting Information
files.

Currently, it is discussed whether anesthetics cause an alteration of the epigenetic landscape
of the cell, which might induce a long-lasting cognitive impairment [5]. One common postop-
erative complication in elderly critically ill patients is the postoperative delirium (POD) that is
also associated with a worse outcome, longer stay on the intensive care unit and higher health-
care related costs [6]. In addition, delirium is also linked to an increased risk of long term cog-
nitive impairments that recover with high inter-individual differences from days to months
[7]. Especially the use of benzodiazepine is, in addition to blood transfusion, one of the only
modifiable factors with strong evidence for an association with delirium after surgery [8].
Within the group of benzodiazepines midazolam shows highest incidence of POD [9].
Although there are some theories that could explain the positive correlation between mida-
zolam administration and the high incidence of POD, such as the degree of sedation [8] and
the function of midazolam as a GABAergic agent [10], the underlying molecular mechanisms
and the pathogenesis of POD still remain elusive.
Currently, a pathogenesis is discussed involving a reduced concentration of the neurotrans-
mitter acetylcholine [11], neuroinflammation [12,13] or decreased antiinflammation [14].
The hydrolysis of acetylcholine is mainly mediated by acetylcholinesterase (AChE) and butyr-
ylcholinesterase (BChE) that can be found in the brain, red blood cells, and central nervous
system [15]. Especially an altered activity and concentration of BChE seems to impact patho-
genesis of POD [16–18] and BChE activity also shows high prognostic capability for POD [19].
Recently, we could demonstrate that the GABAergic agent propofol changes the epigenome
[20]. In context with POD and anesthesia, the expression of lysine-specific demethylase
(KDM1A) seems of special interest as it is associated with cognitive function [21] and demeth-
ylates histone 3 lysine 4 [22]. Hence long-lasting effects on the central nervous system and cog-
nitive abilities caused by the GABAergic midazolam could be caused by changing the
epigenetic landscape of the cells [23–25]. Since it is currently unknown whether and how mid-
azolam influences the activity of the ACHE or BCHE gene. However, we speculate that one
possible mechanism is the alteration of the expression of cholinergic genes by changing the
epigenetic profile of the cells.
Therefore, in this study we investigated whether the expression, activity and methylation
profile of cholinesterases are changed by midazolam. Furthermore, we study whether midazo-
lam changes the epigenetic landscape of the cell by altering KDM1A expression.
2. Materials and methods
2.1. Cell culture
Human neuroblastoma cells SH-SY5Y and the glioblastoma cell line U343 (origin: Cell Lines
Service, CLS, Eppelheim Germany, SH-SY5Y item number: 300154 and U343 item number:
300365) were cultured in Dulbecco’s modified Eagle medium (DMEM; Gibco, Darmstadt,
Germany) at 37˚C and 5% CO
2
with 10% fetal calf serum (FCS; Gibco, Darmstadt, Germany)
and 1% penicillin/streptomycin (Penstrep; Gibco, Darmstadt, Germany). Cells were main-
tained every three to four days by adding 5 ml of Trypsin-EDTA 0.25% (Gibco, Darmstadt,
Germany) after medium removal to dissolve adhesive cells. Furthermore, peripheral blood
mononuclear cells (PBMCs) were examined, after the Ethics Committee’s approval (Ethics
Committee of the Ruhr-University Bochum, Bochum, Germany; ref: 17-5964-BR), registration
at the German Clinical Trials Register (ref: DRKS00012961, https://www.drks.de/drks_web/
navigate.do?navigationId=trial.HTML&TRIAL_ID=DRKS00012961) and written informed
consent. 80 ml EDTA blood was taken from eight healthy donors (5 female and 3 male) and
PBMCs were isolated, using density gradient centrifugation with Ficoll-Paque (GE Healthcare,
Chalfont, UK).
PLOS ONE
Midazolam and epigenetic changes in BCHE
PLOS ONE | https://doi.org/10.1371/journal.pone.0271119 July 8, 2022 2 / 12
Competing interests: The authors have declared
that no competing interests exist.

2.2. Quantitative reverse transcription PCR
q-RT-PCR on SH-SY5Y cells, U343 and PBMCs was performed as described previously [26].
Briefly, cells were cultured in 6-well culture plates and incubated with 250 ng/ml, 1 μg/ml or
50 μg/ml midazolam (midazolam hydrochloride injection solution, B. Braun Melsungen) for
2, 4 and 24 h, 10 μg/ml and 50 μg/ml flumazenil or were left untreated (control). Flumazenil
incubation was performed two hours after starting midazolam incubation. For incubation, the
highest concentration of midazolam (SH-SY5Y 50 μg/ml; U343 250 ng/ml; BV-2 10 μg/ml)
and flumazenil was used, which did not reduce cell viability in different cell lines in previous
experiments. Cells were incubated at 37˚C and 5% CO
2
. After RNA isolation and cDNA syn-
thesis of 1 μg RNA using the QuantiTect Reverse Transcription kit (Qiagen, Hilden, Ger-
many), we utilized 2.5 μl of cDNA together with specific primers (Table 1) and GoTaq qPCR
master mix (Promega, Madison, WI, USA) for a standard qPCR reaction protocol.
2.3. Cholinesterase activity after incubation with midazolam
Cholinesterase activity in SH-SY5Y cells was measured after stimulation with midazolam.
For this purpose, 5 x 105 SH-SY5Y cells were seeded in 4 ml of growth medium containing
10% FBS. Cells were incubated for 24 h at 37˚C and incubated for 2, 4 and 24 h with 50 μg/ml
midazolam or were left untreated.
The proteins were isolated as previously described [20] after washing the cells with PBS.
After the lysates were collected from all experiments, protein quantification was performed
using the Rotiquant universal kit (Roth, Karlsruhe, Germany). The lysates were used for detec-
tion of cholinesterase activity using an acetylcholinesterase assay kit (fluorometric red)
(Abcam, Cambridge, UK) according to the manufacturer’s instructions.
Table 1. Primer pairs for PCR.
Primer name Sequence (5’ to 3’) Product size (bp)
BCHE_M1_SE ATTTAGGTTAAAACGGTGAAATTTC 172
BCHE_M1_AS AAACTAAAATACCGTAACGCGAT
BCHE_U1_SE TTAGGTTAAAATGGTGAAATTTTGG 173
BCHE_U1_AS CTCAAACTAAAATACCATAACACAAT
ACHE_M_SE1 AAT TTT ATT AGT TTC GAG CGA GAT C 189
ACHE_M_AS1 GAC CCA AAA ACC TAC AAC GAC
ACHE_U_SE1 TTT TAT TAG TTT TGA GTG AGA TTG A 188
ACHE_U_AS1 CAA CCC AAA AAC CTA CAA CAA C
ACTB_SE CTGGAACGGTGAAGGTGACA 140
ACTB_AS AAGGGACTTCCTGTAACAATGCA
KDM1A_RT_SE GCCCACTTTATGAAGCCAACG 161
KDM1A_RT_AS GCCAAGGGACACAGGCTTAT
ACHE_mRNA_SE GCT TCA GCA AAG ACA ACG AG 115
ACHE_mRNA_AS GTG TAA TGC AGG ACC ACA GC
BCHE_mRNA_SE ATCCTGCATTTCCCCGAAGT 239
BCHE_mRNA_AS CCGTGCCACCAAAAACTGTC
BCHE_Prom_SE GCATGTGCACTGCAAGTTGA 90
AACTCTCGCGAGCTTTGTCA
BCHE_Prom_AS CCCTGCAGGCAGTCATACAT
CTGCTGCTCCAGCCTGTAAA
https://doi.org/10.1371/journal.pone.0271119.t001
PLOS ONE
Midazolam and epigenetic changes in BCHE
PLOS ONE | https://doi.org/10.1371/journal.pone.0271119 July 8, 2022 3 / 12

2.4. Methylation and expression of BCHE gene after incubation with
midazolam
The DNA methylation of BCHE gene was quantified using methylation-specific PCR after bisul-
phite conversion in SH-SY5Y cells, before and after incubation. For this purpose, 5 x 105 SH-SY5Y
cells per 4 ml were seeded in 6-well culture plates and incubated for 24 h at 37˚ C and 5% CO2.
The cells were incubated with 50 μg/ml or 250 ng/ml midazolam depending on cell type for 2, 4
and 24h. Subsequently, the DNA was isolated using the QIAamp DNA blood mini kit (Qiagen, Hil-
den, Germany), following the manufacturer’s instructions. Bisulphite conversion was performed
with the EZ DNA methylation-gold kit (Zymo Research, Irvine, CA, USA). All DNA samples were
diluted to 10 ng /μl qPCR was performed to detect methylation, as previously described [27], with
the GoTaq qPCR master mix (Promega, Madison, WI, USA) and specific primers (Table 1).
The percentage of methylation was analyzed as previously described [27,28].
2.5. Analysis of histone modifications
Furthermore, histone modifications of histone 3 after incubation were analyzed. SH-SY5Y
cells and U343 were seeded, as previously described, and incubated with 250 ng / ml of mida-
zolam or left untreated (control) for 24 h exactly as previously described [20].
Histone concentration was quantified by the Rotiquant universal kit (Roth, Karlsruhe, Ger-
many) and histone modification was quantified by ELISA using 50 ng protein for the PathS-
can1Di-Methyl-Histone H3 (Lys4) Sandwich ELISA kit (Cell Signaling Technology,
Cambridge, UK).
2.6. Chromatin immunoprecipitation assay (ChIP assay)
A ChIP assay was used to analyze if the promoter of the cholinergic gene BCHE binds to his-
tone H3 lysine K4., 1 x 10
6
SH-SY5Y were used for the Pierce agarose Chip kit (Thermo Fisher
Scientific, Waltham, MA, USA). The H3K4me
2
polyclonal antibody (EpiGentek, Farmingdale,
NY, USA) was used as a specific antibody. As a positive control, an antibody against RNA
polymerase II in combination with specific primers against GAPDH was used, while Rabbit
IgG in combination with our primers against BCHE gene regions was used as a negative con-
trol. After DNA isolation, PCR (One Taq Master Mix, New England Biolabs, Frankfurt am
Main, Germany) was carried out with, BCHE_prom primers (Table 1), and the PCR products
were analyzed on agarose gel (Peqlab, Erlangen, Germany).
2.7. LegendPlex assay for the quantification of cytokines (TNFαand IL6)
To measure cytokine release from glial cells, BV2 cells (kind gift from Veselin Grozdanov
Department of Neurology, Ulm University, Ulm, Germany) were used. Cell culture superna-
tant was utilized after midazolam and LPS treatment for quantification of TNFα, IL6 with the
Legend Plex InflammationPanel (BioLegend, San Diego, CA), according to manufacturer’s
recommendations. Briefly, cells were treated with 1 μg/mL midazolam 100 ng/ml LPS or left
untreated and incubated for 2, 4 and 24 h in complete growth medium. Cell supernatant was
collected and stored at -80˚C until use for cytokine quantification. Measurement was per-
formed using FACS Canto II (Becton Dickinson GmbH, Heidelberg, Germany) according to
the manufacturer’s instructions and analysis was performed using LEGENDplex v8.0 software.
2.8. Statistics
All experiments were performed in duplicate and repeated at least three times. Results are pre-
sented as mean ±standard deviation. If not otherwise stated, all datasets were analyzed using
PLOS ONE
Midazolam and epigenetic changes in BCHE
PLOS ONE | https://doi.org/10.1371/journal.pone.0271119 July 8, 2022 4 / 12

an unpaired t-Test or one-way ANOVA for multiple comparisons with a Dunnett’s multiple
comparisons test for specific comparisons. A p-value �0.05 was considered statistically signif-
icant. For multiple comparisons, specific comparisons were only analyzed if the one-way
ANOVA showed a statistically significant difference between the groups. All statistical analyses
were performed using GraphPad Prism 8 (San Diego, CA, USA).
3. Results
3.1. The activity and expression of AChE and BChE are altered after
incubation with midazolam
Cholinesterase activity in SH-SY5Y cells was gradually reduced after incubation with midazo-
lam, without reaching statistical significance (control: mean + SD 90.7 +5.5; 2 hours mean
+ SD 42.51 + 28.7; p = n.s.; Fig 1A) but the intracellular AChE and BChE activity (p = 0.01; Fig
1A) and ACHE (p<0.01) and BCHE (p = 0.03) mRNA expression increased after 24 h by
about 80% (Fig 1B and 1C).
3.2. Application of the midazolam antagonist flumazenil reverses
midazolam induced effects on BCHE expression
In order to elucidate if midazolam antagonist flumazenil is capable to reduce midazolam
induced effects on ACHE and BCHE expression, SH-SY5Y cells were incubated with flumaze-
nil two hours after midazolam exposure. Here, we show a gradual increasing abolition of the
midazolam effect (increased BCHE expression) under increasing doses of the antagonist flu-
mazenil (Fig 2B). After the addition of 10μg/mL flumazenil the increased expression associated
with midazolam of BCHE was reduced (p<0.05; Fig 2B). Interestingly, this effect was not
observed on ACHE expression (Fig 2A).
3.3. Midazolam induces epigenetic changes in the BCHE gene of neuronal
cells
Midazolam induced a decrease in BCHE intron 2 DNA methylation (p = 0.01; Fig 3A) and in
the di-methylation of H3K4 (p = 0.02; Fig 3B), where BCHE promoter binds (Fig 3C). ACHE
DNA methylation was not altered by incubation with 50 μg/ml midazolam (Fig 3D).
Fig 1. Activity and mRNA expression of AChE and BChE after incubation with midazolam A) the intracellular cholinesterase activity increased
24 h after midazolam exposure (n = 4; p = 0.01) and was measured by fluorometric assay B) ACHE mRNA quantified by qPCR expression was
increased 24 hours after midazolam exposure in SH-SY5Y cells (n = 3; p<0.01) C) BCHE mRNA quantified by qPCR expression was increased
24 hours after midazolam exposure in SH-SY5Y cells (n = 3; p = 0.03). Data are presented as mean ±standard deviation. The reported p-value
refers to the Dunnett’s post-hoc test, comparing the underlying columns at the ends of each bar.
https://doi.org/10.1371/journal.pone.0271119.g001
PLOS ONE
Midazolam and epigenetic changes in BCHE
PLOS ONE | https://doi.org/10.1371/journal.pone.0271119 July 8, 2022 5 / 12

Fig 3. Methylation of BCHE in neuronal SH-SY5Y and U343 cells after midazolam incubation A) BCHE intron 2
methylation reduced after midazolam (50 μg/ml) exposure of SH-SY5Y cells (n = 3; p = 0.01) analyzed by methylation
specific PCR. B) ELISA showed that histone H3 lysine 4 di-methylation (H3K4me2) decreased in U343 cells after
incubation with 250 ng/ml midazolam (n = 3; p = 0.02) C) Chip-Assay confirmed binding of BCHE promoter region
(90 bp) to H3K4me2; a 100 bp DNA Ladder was utilized; lanes 1, 7 show incubation with H3K27 antibody; lanes 2 and
6 show incubation with H3K4 antibody; lane 4 shows negative control without antibody and lane 5 shows positive
control with RNA-polymerase II antibody (two experiments out of three are shown; n = 3). D) ACHE -571-/-670
promoter methylation was not affected by midazolam (50 μg/ml) exposure of SH-SY5Y cells (n = 3; p = n.s.) analyzed
by methylation specific PCR. Data are presented as mean ±standard deviation.
https://doi.org/10.1371/journal.pone.0271119.g003
Fig 2. SH-SY5Y cells were incubated with midazolam for 24 h and with flumazenil (starting 2 h after midazolam
exposure) for 22 hours. ACHE and BCHE mRNA expression relative to β-Actin mRNA expression were quantified by
qPCR. Incubation with flumazenil A) did not alter ACHE expression (n = 6; p = n.s.) and reduced B) BCHE (n = 6;
p = 0.046) expression. Data are presented as mean ±standard deviation. The reported p-value refers to the Dunnett’s
post-hoc test, comparing the underlying columns at the ends of each bar.
https://doi.org/10.1371/journal.pone.0271119.g002
PLOS ONE
Midazolam and epigenetic changes in BCHE
PLOS ONE | https://doi.org/10.1371/journal.pone.0271119 July 8, 2022 6 / 12

3.4. Midazolam increases the expression of lysine specific demethylase
KDM1A
To explore the underlying mechanisms for the decrease in H3K4me2, we analyzed the expres-
sion of lysine specific demethylase KDM1A after exposure to midazolam. KDM1A mRNA
expression was increased in, in U343 by about 50% (p <0.01; Fig 4A), in PBMCs by more than
100% (p<0.01; Fig 4B) and in SH-SY5Y by about 50% (p = 0.0038; Fig 4C). Incubation with
flumazenil reduced midazolam induced effects in a visible dose dependent manner in
SH-SY5Y cells, while incubation with midazolam alone led to increased expression of KDM1A
(p <0.001; Fig 4D) expression.
3.5. Midazolam does not induce the release of cytokines from BV-2 glial cells
Since postoperative delirium is strongly associated with neuroinflammation, we finally investi-
gated whether midazolam itself evoked cytokine secretion in neural glial cells (BV-2, RRID:
CVCL_0182). Midazolam did not induce any change in cytokine secretion in BV-2 cells
(p = ns), compared to untreated cells. Cells incubated with lipopolysaccharide (LPS) as positive
control had higher TNF-αcytokine levels (p = 0.01; Fig 5A) and higher IL-6 levels (p = 0.02;
Fig 5B) compared to cells incubated with midazolam for 24 hours.
4. Discussion
Midazolam is a widely used benzodiazepine although its application is associated with the
occurrence of POD [9]. A potential mechanism for the development of delirium is impaired
cholinergic transmission based on the deficiency of acetylcholine in the brain [29]. However,
as the causal relationship between midazolam and the cholinergic system is unknown, we sys-
tematically analyzed the expression and epigenetic regulation of cholinergic genes in neuronal
Fig 4. KDM1A mRNA expression was quantified relative to β-Actin mRNA expression by qPCR. Increased
expression of lysine specific demethylase (KDM1A) in different cells after midazolam [50 μg/ml] exposure for 24 hours
analyzed by qPCR. KDM1A expression increased in U343 (n = 3; A, in peripheral blood mononuclear cells (PBMCs)
(n = 8; B) and in SH-SY5Y (n = 3; C). Flumazenil did not reduceKDM1A expression (n = 6, D). Data are presented as
mean ±standard deviation. The reported p-value refers to the Dunnett’s post-hoc test, comparing the underlying
columns at the ends of each bar.
https://doi.org/10.1371/journal.pone.0271119.g004
PLOS ONE
Midazolam and epigenetic changes in BCHE
PLOS ONE | https://doi.org/10.1371/journal.pone.0271119 July 8, 2022 7 / 12

cells after midazolam exposure. As a different postoperative activity of the proteins AChE and
BChE in patients is already described [16,30,31] it seems of special interest, how their gene
expression is regulated after midazolam exposure.
First, we could detect a visibly early decrease in cholinesterase activity and a slight decrease
in the expression of BCHE mRNA, but a late increase in the activity and the expression of
ACHE and BCHE mRNA. Our results regarding AChE and BChE activity and expression are
in line with other studies analyzing AChE and BChE activities in peripheral blood from preop-
erative and postoperative patients [16,19,30]. AChE and BChE concentrations in blood and
cerebrospinal fluid were altered in patients undergoing total hip/knee replacement, and BChE
concentration showed the highest prognostic value for the development of POD [19]. Thus,
increased gene expression, especially BChE, could represent an important mechanism, as it
could be found in the brain of patients with Alzheimer’s disease [32] and several studies
explored the therapeutic implication of cholinesterase inhibitors in alleviating postoperative
delirium [33].
Second, we tested the methylation of a BCHE gene region and a histon, with BCHE binding
affinity. Methylation of the BCHE gene region (Intron 2) and the H3K4 di-methylation
decreased after midazolam incubation. Thus, it seems appropriate to suggest that the region of
the BCHE gene we investigated has activating effects on the transcription of this gene. How-
ever, it must be mentioned that the reduction in methylation was only about 10%. This seems
to be questionable for a more than doubled amount of mRNA expression. In fact, other studies
have already shown that a small change in DNA methylation of approximately 5% can have a
great impact on gene expression [34]. Therefore, it seems possible that this small change in
methylation state may cause this effect on mRNA expression.
Third: Since midazolam changed the di-methylation of H3K4, and we could detect binding
of BCHE to this histone, it seems appropriate that midazolam might change the epigenome of
the cell by influencing histon-modifying enzymes. H3K4me2 has been shown to mark actively
transcribing genes [35]. In our analyzed neuronal cell line di-methylation was nearly 100 per-
cent and midazolam could decrease the methylation slightly. A reduction of the di-methylation
of H3K4 could therefore mean an overall increase in BChE expression. The demethylation of
H3K4 is facilitated by KDM1A and is a well-established mechanism underlying transcriptional
gene repression, but recently its role in gene activation could be shown [36]. The KDM1A
demethylation of H3K4me2 in GR-targeted enhancers was shown to be important for GC-
Fig 5. Cytokine secretion in BV-2 glial cells after midazolam (1μg/ml) for 24 hours and LPS (100 ng/ml) for 4
hours (n = 3). BV-2 cells were incubated with midazolam (1μg/ml) or lipopolysaccharide LPS (100 ng/ml) or left
untreated. Cytokine expression was quantified using a bead-based immunoassay. Data are presented as
mean ±standard deviation. The reported p-value refers to the Dunnett’s post-hoc test, comparing the underlying
columns at the ends of each bar.
https://doi.org/10.1371/journal.pone.0271119.g005
PLOS ONE
Midazolam and epigenetic changes in BCHE
PLOS ONE | https://doi.org/10.1371/journal.pone.0271119 July 8, 2022 8 / 12

mediated gene transcription, facilitating a molecular mechanism for the demethylation of
H3K4me2 in gene activation [36]. Since changes in the methylation of histone 3 is facilitated
by KDM1A, we analyzed the expression of KDM1A in our cell lines, and because POD is also
associated with an altered cholinesterase activity in blood samples [16], we additionally investi-
gated the expression of these enzymes in PBMCs. Strikingly, KDM1A showed a significant
increased expression after midazolam treatment in all investigated cell lines (including
PBMCs). Thus, our results provide first evidence that midazolam indeed rewrites the epige-
netic landscape of the cell. Interestingly, the application of KDM1A inhibitors is associated
with positive effects on memory. Recently it could be demonstrated that inhibition of KDM1A
corrects memory deficit and behavior alterations in a mouse model of Alzheimer’s Disease
[21]. Another KDM1A inhibitor T-448 improved learning function in mice suffering from
neuronal glutamate receptor hypofunction [37]. Thus, it seems tempting to speculate that
KDM1A inhibitors might represent a therapeutic approach against POD. However, this crude
thesis needs to be evaluated in upcoming studies.
Fourth: As increased expression of BCHE seems to be critical mechanisms after midazolam
exposure. In this context, we analyzed if the midazolam antagonist flumazenil could inhibit
midazolam induced effects. Indeed, we could show that flumazenil application reduced mida-
zolam-induced expression in a dose-dependent manner. Regarding the effects of flumazenil
application after midazolam anesthesia on brain function, we can only speculate. However, it
is known that cognitive abnormalities can significantly be ameliorated after benzodiazepine
use by slow subcutaneous infusion of flumazenil [38] and that flumazenil administration atten-
uates cognitive impairment [38]. Therefore, flumazenil use might be effective in reducing
POD.
Lastly, since POD is related to neuroinflammation [39], we analyzed if there is a link
between midazolam treatment for neuroinflammation. We could demonstrate that in our pos-
itive control, the incubation of glial cells with LPS TNF-alpha and IL-6 were significantly upre-
gulated. However, midazolam treatment had no influence on the expression of these
cytokines. IL-6 seems to be of particular interest as it seems to be a consistent predictor of
delirium in surgical samples [40]. Therefore, we can conclude that midazolam does not
strongly contribute to pro-inflammatory signaling, being discussed as additional factors in the
development of POD [12–14].
We have to discuss the limitations of our study. Direct transfer to the bedside is inappropri-
ate because we worked with cell lines as a model for the human brain. However, for instance
we chose the neuronal cell line SH-SY5Y, because these represent an established cell line used
to study brain disorders such as Alzheimer’s disease or Parkinson [41,42]. In addition, the
extraction of neuronal cells from healthy volunteers or patients with POD is ethically not feasi-
ble [43]. Despite great efforts made to achieve the highest possible degree of standardization,
variance in effect sizes or observed effects can occur within the individual experiments, which
limits the statistical or mathematical accuracy of our experiments. However, this had no or
only a negligible effect on the interpretation of our data. Therefore, considering the limitations
of immortalized cell lines, we are confident that it is appropriate to perform our investigations
in our selected cell lines. In addition, direct measurement of acetylcholine would be interesting
but is not suitable as it is extremely unstable [44]. Thus, we mainly refer to the central effectors
and regulators of acetylcholine concentration.
5. Conclusions
In summary, we found that midazolam upregulates intracellular BCHE expression. This upre-
gulation in expression might be caused by demethylation of BCHE gene and H3K4 me2
PLOS ONE
Midazolam and epigenetic changes in BCHE
PLOS ONE | https://doi.org/10.1371/journal.pone.0271119 July 8, 2022 9 / 12

demethylation and be facilitated by KDM1A. Thus, our results underpin the thesis, that over-
expression of BCHE might aggravate postoperative delirium, due to an increased hydrolysis of
acetyl-choline. Although POD is closely related to neuroinflammation, midazolam appears to
be a separate trigger, independent of inflammation. Further studies should validate our prom-
ising results and mechanistic implications in the clinical context regarding feasibility and
transferability.
Supporting information
S1 Raw images.
(JPG)
Author Contributions
Conceptualization: Katharina Rump, Hartmuth Nowak.
Data curation: Katharina Rump, Caroline Holtkamp, Tim Rahmel.
Formal analysis: Katharina Rump, Lars Bergmann, Jennifer Orlowski, Patrick Thon, Bjo¨rn
Koos.
Investigation: Katharina Rump, Caroline Holtkamp, Matthias Unterberg.
Methodology: Katharina Rump, Caroline Holtkamp, Hartmuth Nowak, Matthias Unterberg,
Jennifer Orlowski, Patrick Thon, Zainab Bazzi, Maha Bazzi, Tim Rahmel.
Project administration: Michael Adamzik, Bjo¨rn Koos, Tim Rahmel.
Supervision: Lars Bergmann, Hartmuth Nowak, Michael Adamzik, Tim Rahmel.
Writing – original draft: Katharina Rump.
Writing – review & editing: Caroline Holtkamp, Lars Bergmann, Hartmuth Nowak, Matthias
Unterberg, Michael Adamzik, Bjo¨rn Koos, Tim Rahmel.
References
1. Halbeck E, Dumps C, Bolkenius D. [Drugs for intravenous induction of anesthesia: ketamine, midazo-
lam and synopsis of current hypnotics]. Anaesthesist. 2018; 67(8):617–34. https://doi.org/10.1007/
s00101-018-0469-7 PMID: 30069734
2. Khanderia U, Pandit SK. Use of midazolam hydrochloride in anesthesia. Clin Pharm. 1987; 6(7):533–
47. PMID: 3319363
3. Alwardt CM, Redford D, Larson DF. General anesthesia in cardiac surgery: a review of drugs and prac-
tices. J Extra Corpor Technol. 2005; 37(2):227–35. PMID: 16117465
4. Li WX, Luo RY, Chen C, Li X, Ao JS, Liu Y, et al. Effects of propofol, dexmedetomidine, and midazolam
on postoperative cognitive dysfunction in elderly patients: a randomized controlled preliminary trial.
Chin Med J (Engl). 2019; 132(4):437–45.
5. Ji M, Dong L, Jia M, Liu W, Zhang M, Ju L, et al. Epigenetic enhancement of brain-derived neurotrophic
factor signaling pathway improves cognitive impairments induced by isoflurane exposure in aged rats.
Mol Neurobiol. 2014; 50(3):937–44. https://doi.org/10.1007/s12035-014-8659-z PMID: 24553857
6. Titsworth WL, Hester J, Correia T, Reed R, Guin P, Archibald L, et al. The effect of increased mobility
on morbidity in the neurointensive care unit. J Neurosurg. 2012; 116(6):1379–88. https://doi.org/10.
3171/2012.2.JNS111881 PMID: 22462507
7. Silverstein JH, Timberger M, Reich DL, Uysal S. Central nervous system dysfunction after noncardiac
surgery and anesthesia in the elderly. Anesthesiology. 2007; 106(3):622–8. https://doi.org/10.1097/
00000542-200703000-00026 PMID: 17325520
8. Devlin JW, Skrobik Y, Ge
´linas C, Needham DM, Slooter AJC, Pandharipande PP, et al. Clinical Practice
Guidelines for the Prevention and Management of Pain, Agitation/Sedation, Delirium, Immobility, and
PLOS ONE
Midazolam and epigenetic changes in BCHE
PLOS ONE | https://doi.org/10.1371/journal.pone.0271119 July 8, 2022 10 / 12

Sleep Disruption in Adult Patients in the ICU. Crit Care Med. 2018; 46(9):e825–e73. https://doi.org/10.
1097/CCM.0000000000003299 PMID: 30113379
9. Cui Y, Li G, Cao R, Luan L, Kla KM. The effect of perioperative anesthetics for prevention of postopera-
tive delirium on general anesthesia: A network meta-analysis. J Clin Anesth. 2020; 59:89–98. https://
doi.org/10.1016/j.jclinane.2019.06.028 PMID: 31284222
10. Zhou Y, Li Y, Wang K. Bispectral Index Monitoring During Anesthesia Promotes Early Postoperative
Recovery of Cognitive Function and Reduces Acute Delirium in Elderly Patients with Colon Carcinoma:
A Prospective Controlled Study using the Attention Network Test. Med Sci Monit. 2018; 24:7785–93.
https://doi.org/10.12659/MSM.910124 PMID: 30378591
11. Zujalovic B, Barth E. Delirium Accompanied by Cholinergic Deficiency and Organ Failure in a 73-Year-
Old Critically Ill Patient: Physostigmine as a Therapeutic Option. Case Rep Crit Care. 2015;
2015:793015. https://doi.org/10.1155/2015/793015 PMID: 26550498
12. Cascella M, Muzio MR, Bimonte S, Cuomo A, Jakobsson JG. Postoperative delirium and postoperative
cognitive dysfunction: updates in pathophysiology, potential translational approaches to clinical practice
and further research perspectives. Minerva Anestesiol. 2018; 84(2):246–60. https://doi.org/10.23736/
S0375-9393.17.12146-2 PMID: 28984099
13. Umholtz M, Nader ND. Anesthetic Immunomodulation of the Neuroinflammation in Postoperative Cog-
nitive Dysfunction. Immunol Invest. 2017; 46(8):805–15. https://doi.org/10.1080/08820139.2017.
1373898 PMID: 29058541
14. Maldonado JR. Neuropathogenesis of delirium: review of current etiologic theories and common path-
ways. Am J Geriatr Psychiatry. 2013; 21(12):1190–222. https://doi.org/10.1016/j.jagp.2013.09.005
PMID: 24206937
15. Das UN. Acetylcholinesterase and butyrylcholinesterase as markers of low-grade systemic inflamma-
tion. Ann Hepatol. 2012; 11(3):409–11. PMID: 22481463
16. Mu¨ller A, Olbert M, Heymann A, Zahn PK, Plaschke K, von Dossow V, et al. Relevance of peripheral
cholinesterase activity on postoperative delirium in adult surgical patients (CESARO): A prospective
observational cohort study. Eur J Anaesthesiol. 2019; 36(2):114–22. https://doi.org/10.1097/EJA.
0000000000000888 PMID: 30431498
17. John M, Ely EW, Halfkann D, Schoen J, Sedemund-Adib B, Klotz S, et al. Acetylcholinesterase and
butyrylcholinesterase in cardiosurgical patients with postoperative delirium. J Intensive Care. 2017;
5:29. https://doi.org/10.1186/s40560-017-0224-1 PMID: 28560042
18. Adam EH, Haas V, Lindau S, Zacharowski K, Scheller B. Cholinesterase alterations in delirium after
cardiosurgery: a German monocentric prospective study. BMJ Open. 2020; 10(1):e031212. https://doi.
org/10.1136/bmjopen-2019-031212 PMID: 31941763
19. Lin X, Tang J, Liu C, Li X, Cao X, Wang B, et al. Cerebrospinal fluid cholinergic biomarkers are associ-
ated with postoperative delirium in elderly patients undergoing Total hip/knee replacement: a prospec-
tive cohort study. BMC Anesthesiol. 2020; 20(1):246. https://doi.org/10.1186/s12871-020-01166-9
PMID: 32988381
20. Holtkamp C, Koos B, Unterberg M, Rahmel T, Bergmann L, Bazzi Z, et al. A novel understanding of
postoperative complications: In vitro study of the impact of propofol on epigenetic modifications in cho-
linergic genes. PLoS One. 2019; 14(5):e0217269. https://doi.org/10.1371/journal.pone.0217269 PMID:
31141559
21. Maes T, Mascaro
´C, Rotllant D, Lufino MMP, Estiarte A, Guibourt N, et al. Modulation of KDM1A with
vafidemstat rescues memory deficit and behavioral alterations. PLoS One. 2020; 15(5):e0233468.
https://doi.org/10.1371/journal.pone.0233468 PMID: 32469975
22. Rudolph T, Beuch S, Reuter G. Lysine-specific histone demethylase LSD1 and the dynamic control of
chromatin. Biol Chem. 2013; 394(8):1019–28. https://doi.org/10.1515/hsz-2013-0119 PMID: 23612539
23. Culley DJ, Baxter MG, Yukhananov R, Crosby G. Long-term impairment of acquisition of a spatial mem-
ory task following isoflurane-nitrous oxide anesthesia in rats. Anesthesiology. 2004; 100(2):309–14.
https://doi.org/10.1097/00000542-200402000-00020 PMID: 14739805
24. Culley DJ, Baxter M, Yukhananov R, Crosby G. The memory effects of general anesthesia persist for
weeks in young and aged rats. Anesth Analg. 2003; 96(4):1004–9. https://doi.org/10.1213/01.ANE.
0000052712.67573.12 PMID: 12651650
25. Mori K, Iijima N, Higo S, Aikawa S, Matsuo I, Takumi K, et al. Epigenetic suppression of mouse Per2
expression in the suprachiasmatic nucleus by the inhalational anesthetic, sevoflurane. PLoS One.
2014; 9(1):e87319. https://doi.org/10.1371/journal.pone.0087319 PMID: 24498074
26. Rump K, Brendt P, Frey UH, Schafer ST, Siffert W, Peters J, et al. Aquaporin 1 and 5 expression
evoked by the beta2 adrenoreceptor agonist terbutaline and lipopolysaccharide in mice and in the
human monocytic cell line THP-1 is differentially regulated. Shock. 2013; 40(5):430–6. https://doi.org/
10.1097/SHK.0000000000000035 PMID: 24088990
PLOS ONE
Midazolam and epigenetic changes in BCHE
PLOS ONE | https://doi.org/10.1371/journal.pone.0271119 July 8, 2022 11 / 12

27. Unterberg M, Kreuzer MJ, Schafer ST, Bazzi Z, Adamzik M, Rump K. NFKB1 Promoter DNA from nt
+402 to nt+99 Is Hypomethylated in Different Human Immune Cells. PLoS One. 2016; 11(6):e0156702.
https://doi.org/10.1371/journal.pone.0156702 PMID: 27249028
28. Agrawal S, Unterberg M, Koschmieder S, zur Stadt U, Brunnberg U, Verbeek W, et al. DNA methylation
of tumor suppressor genes in clinical remission predicts the relapse risk in acute myeloid leukemia.
Cancer Res. 2007; 67(3):1370–7. https://doi.org/10.1158/0008-5472.CAN-06-1681 PMID: 17283175
29. Hshieh TT, Fong TG, Marcantonio ER, Inouye SK. Cholinergic deficiency hypothesis in delirium: a syn-
thesis of current evidence. J Gerontol A Biol Sci Med Sci. 2008; 63(7):764–72. https://doi.org/10.1093/
gerona/63.7.764 PMID: 18693233
30. Cerejeira J, Batista P, Nogueira V, Firmino H, Vaz-Serra A, Mukaetova-Ladinska EB. Low preoperative
plasma cholinesterase activity as a risk marker of postoperative delirium in elderly patients. Age Ageing.
2011; 40(5):621–6. https://doi.org/10.1093/ageing/afr053 PMID: 21576115
31. Cerejeira J, Nogueira V, Luı
´s P, Vaz-Serra A, Mukaetova-Ladinska EB. The cholinergic system and
inflammation: common pathways in delirium pathophysiology. J Am Geriatr Soc. 2012; 60(4):669–75.
https://doi.org/10.1111/j.1532-5415.2011.03883.x PMID: 22316182
32. Bono GF, Simão-Silva DP, Batistela MS, Josviak ND, Dias PF, Nascimento GA, et al. Butyrylcholines-
terase: K variant, plasma activity, molecular forms and rivastigmine treatment in Alzheimer’s disease in
a Southern Brazilian population. Neurochem Int. 2015; 81:57–62. https://doi.org/10.1016/j.neuint.2014.
12.009 PMID: 25624079
33. Yu A, Wu S, Zhang Z, Dening T, Zhao S, Pinner G, et al. Cholinesterase inhibitors for the treatment of
delirium in non-ICU settings. Cochrane Database Syst Rev. 2018; 6(6):Cd012494. https://doi.org/10.
1002/14651858.CD012494.pub2 PMID: 29952000
34. Butts B, Butler J, Dunbar SB, Corwin E, Gary RA. Effects of Exercise on ASC Methylation and IL-1
Cytokines in Heart Failure. Med Sci Sports Exerc. 2018; 50(9):1757–66. https://doi.org/10.1249/MSS.
0000000000001641 PMID: 29683921
35. Hyun K, Jeon J, Park K, Kim J. Writing, erasing and reading histone lysine methylations. Exp Mol Med.
2017; 49(4):e324. https://doi.org/10.1038/emm.2017.11 PMID: 28450737
36. Clark EA, Wu F, Chen Y, Kang P, Kaiser UB, Fang R, et al. GR and LSD1/KDM1A-Targeted Gene Acti-
vation Requires Selective H3K4me2 Demethylation at Enhancers. Cell Rep. 2019; 27(12):3522–32.e3.
https://doi.org/10.1016/j.celrep.2019.05.062 PMID: 31216473
37. Matsuda S, Baba R, Oki H, Morimoto S, Toyofuku M, Igaki S, et al. T-448, a specific inhibitor of LSD1
enzyme activity, improves learning function without causing thrombocytopenia in mice. Neuropsycho-
pharmacology. 2019; 44(8):1505–12. https://doi.org/10.1038/s41386-018-0300-9 PMID: 30580376
38. Federico A, Lugoboni F, Mantovani E, Martini A, Morbioli L, Casari R, et al. Detoxification Improves Mul-
tidomain Cognitive Dysfunction in High-Dose Benzodiazepine Abusers. Front Neurosci. 2020; 14:747.
https://doi.org/10.3389/fnins.2020.00747 PMID: 32848544
39. Yang T, Velagapudi R, Terrando N. Neuroinflammation after surgery: from mechanisms to therapeutic
targets. Nat Immunol. 2020; 21(11):1319–26. https://doi.org/10.1038/s41590-020-00812-1 PMID:
33077953
40. Adamis D, van Gool WA, Eikelenboom P. Consistent patterns in the inconsistent associations of Insu-
lin-like growth factor 1 (IGF-1), C-Reactive Protein (C-RP) and Interleukin 6 (IL-6) levels with delirium in
surgical populations. A systematic review and meta-analysis. Arch Gerontol Geriatr. 2021; 97:104518.
https://doi.org/10.1016/j.archger.2021.104518 PMID: 34536657
41. Inoue N, Ogura S, Kasai A, Nakazawa T, Ikeda K, Higashi S, et al. Knockdown of the mitochondria-
localized protein p13 protects against experimental parkinsonism. EMBO Rep. 2018; 19(3). https://doi.
org/10.15252/embr.201744860 PMID: 29371327
42. Lauzon MA, Faucheux N. A small peptide derived from BMP-9 can increase the effect of bFGF and
NGF on SH-SY5Y cells differentiation. Mol Cell Neurosci. 2018; 88:83–92. https://doi.org/10.1016/j.
mcn.2018.01.003 PMID: 29341901
43. Koskderelioglu A, Onder O, Gucuyener M, Altay T, Kayali C, Gedizlioglu M. Screening for postoperative
delirium in patients with acute hip fracture: Assessment of predictive factors. Geriatr Gerontol Int. 2017;
17(6):919–24. https://doi.org/10.1111/ggi.12806 PMID: 27283273
44. Soreq H, Seidman S. Acetylcholinesterase—new roles for an old actor. Nat Rev Neurosci. 2001; 2
(4):294–302. https://doi.org/10.1038/35067589 PMID: 11283752
PLOS ONE
Midazolam and epigenetic changes in BCHE
PLOS ONE | https://doi.org/10.1371/journal.pone.0271119 July 8, 2022 12 / 12