Forget me not: The effect of doxycycline on human declarative memory

Full text

Forget me not: The effect of doxycycline on human declarative memory
Jelena M Wehrli
a,*
, Yanfang Xia
a
, Laura Meister
a
, Sarrina Tursunova
a
, Birgit Kleim
a
,
Dominik R Bach
a,b
, Boris B Quednow
c,d
a
Department of Adult Psychiatry and Psychotherapy, University Hospital of Psychiatry Zurich, University of Zurich, Lenggstrasse 31, Zurich 8032, Switzerland
b
University of Bonn, Transdisciplinary Research Area "Life and Health", Hertz Chair for Articial Intelligence and Neuroscience, Bonn, Germany
c
Experimental Pharmacopsychology and Psychological Addiction Research, Department of Adult Psychiatry and Psychotherapy, University Hospital of Psychiatry Zurich,
University of Zurich, Zurich, Switzerland
d
Neuroscience Center Zurich, Joint Center of University of Zurich and Swiss Federal Institute of Technology Zurich, Zurich, Switzerland
ARTICLE INFO
Keywords:
Doxycycline
Memory modication
Declarative memory
Neuroprotective
ABSTRACT
Investigations into neuroprotective drugs are in high demand for the treatment of neurodegenerative diseases,
such as multiple sclerosis or Alzheimer’s disease, but also psychiatric disorders, such as depression, trauma, and
substance use. One potential drug class being investigated are tetracyclines impacting on a variety of neuro-
protective mechanisms. At the same time, tetracyclines like doxycycline have been suggested to affect human
fear and spatial memory as well as reducing declarative memory retention. Based on the assumed necessity for
synaptic consolidation in hippocampus-dependent learning, we hypothesised declarative memory may be
similarly impaired by doxycycline as fear and spatial memory. Therefore, in this study we investigate the po-
tential diminishing effects of doxycycline on consolidation of declarative memory in healthy humans. Addi-
tionally, to test for effect specicity we assessed motor memory, sustained attention, and processing speed. We
administered a neuropsychological test battery in three independent randomized placebo-controlled double-
blind trials (RCTs), in which healthy young volunteers (total N =252) either received a single oral dose
doxycycline (200 mg, n =126) or placebo (n =126) in a between-subject design. We found no evidence for a
detrimental effect of doxycycline on declarative memory; instead, doxycycline improved declarative learning (p-
value=0.022, Cohen’s d =0.15) and memory consolidation (p =0.040, d =0.26). Contrarily, doxycycline slightly
reduced motor learning (p =0.001, d =0.10) but subtly strengthened long-term motor memory (p =0.001, d =
0.10). These results suggest that doxycycline can improve declarative learning and memory without having long
term negative effects on other cognitive domains in healthy humans. Our results give hope to further investigate
doxycycline in neuroprotective treatment applications.
1. Introduction
Tetracyclic antibiotics have been clinically used for more than 40
years, due to their antimicrobial properties both outside the central
nervous system (CNS) (Fischer and Ganellin, 2006; Nelson and Levy,
2011) and within the CNS, e.g. to treat neuroborreliosis (Bernardino
et al., 2009; Dotevall et al., 1988; Karlsson et al., 1996). More recently,
tetracyclines such as doxycycline and minocycline have been studied as
potential neuroprotective treatment for neurodegenerative diseases,
such as multiple sclerosis (MS), Alzheimer’s disease (AD), Parkinson’s
disease (PD), and amyotrophic lateral sclerosis (ALS) (Balducci et al.,
2018; Balducci and Forloni, 2019; Bernardino et al., 2009; Bortolanza
et al., 2018; Orsucci et al., 2009; Rahmani et al., 2022; Santa-Cecília
et al., 2019). Various mechanisms including inhibiting microglial acti-
vation (Santa-Cecília et al., 2016; Yrj¨
anheikki et al., 1998), reduction of
apoptosis (Sapadin and Fleischmajer, 2006), and reactive oxygen spe-
cies (ROS) scavenging (Miyachi et al., 1986; Ramamurthy et al., 2009)
are assumed to lead to the neuroprotective effects at the cellular level.
Yet, the tetracyclic antibiotics doxycycline and minocycline are also
inhibitors of the matrix metalloproteinase 9 (MMP9) (Boyle et al., 1998;
Burggraf et al., 2007; Golub et al., 1991; Kim et al., 2005), which is
required for inducing synaptic plasticity necessary for memory consol-
idation (Nagy, 2006; Vafadari et al., 2016). Therefore, it has been sug-
gested that tetracyclines may negatively affect mnemonic processes.
Animal studies have shown reduction in spatial learning and inhibitory
avoidance after MMP9 inhibition (Gross et al., 2022; Nagy et al., 2007;
* Corresponding author.
E-mail address: jelena.wehrli@uzh.ch (J.M. Wehrli).
Contents lists available at ScienceDirect
European Neuropsychopharmacology
journal homepage: www.sciencedirect.com/journal/european-neuropsychopharmacology
https://doi.org/10.1016/j.euroneuro.2024.08.006
Received 8 May 2024; Received in revised form 8 August 2024; Accepted 12 August 2024
European Neuropsychopharmacology 89 (2024) 1–9
Available online 31 August 2024
0924-977X/© 2024 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-
nc-nd/4.0/ ).

Wright et al., 2007). Doxycycline and minocycline have also been tested
in four moderately powered human trials (N =76 to 105: (Bach et al.,
2018, 2019; Wehrli et al., 2023; Xia et al., 2024) assessing fear memory,
of which two indicated consolidation impairment when learning took
place under inuence of a tetracycline (Bach et al., 2018; Xia et al.,
2024). Another small study (N =20) found an impeding effect of min-
ocycline on human spatial memory (Berens et al., 2020).
Declarative memory impairment is an important side-effect to
consider when identifying drugs to be studied for pharmacological
management of neurodegenerative disorders. Based on the memory
impairments detected both in human and animal studies and the
assumed necessity of synaptic consolidation for hippocampus-
dependent declarative memory (Eichenbaum, 2004) we investigate the
potential attenuating effects of doxycycline on declarative memory. As
of today, there is a dearth of studies investigating the effect of tetracyclic
antibiotics on human declarative memory. A small study by Idzikowski
& Oswald (1983) (N =32) observed a decline in memory retention
under doxycycline, but this nding has not yet been replicated.
Here, we address this question by testing the effect of a single oral
dose of 200 mg doxycycline on verbal and visual declarative memory in
a large sample of healthy volunteers (N =252). Additionally, we
employed neuropsychological tests for motor learning, sustained
attention, and processing speed. While motor learning is not dependent
on the hippocampus (Gould et al., 2002) the hippocampus is related to
sequence learning processes (Gheysen et al., 2010) and may be involved
in the consolidation of the motor memory (Albouy et al., 2015; Jaco-
bacci et al., 2020; Schapiro et al., 2019), therefore, we expect no dif-
ferences in the acquisition of the motor memory task, but potential
attenuating effects on the recall. The tests employed to measure sus-
tained attention and processing speed are highly sensitive to changes in
cognitive function (Ogg et al., 2008; Riccio et al., 2002) as well as drug
effects (Greenblatt et al., 2005; Jaeger, 2018; Thapar et al., 1995) and
were employed to test the specicity of effects.
2. Methods
2.1. Aim
In three randomized placebo-controlled double-blind trials (RCTs)
reported in Bach et al. (2019: Experiment 1), Wehrli et al. (2023:
Experiment 2) and Meister et al., (under review: Experiment 3) young
healthy adults received a single oral dose of the antibiotic doxycycline to
test the potential attenuating effects of the drug on fear memory
consolidation and reconsolidation. Experiment 1 employed a Pavlovian
cued fear conditioning task to induce fear, Experiment 2 a Pavlovian
trace fear conditioning task, and Experiment 3 the trauma lm paradigm
(James et al., 2016). In addition, all three studies performed standard-
ized neuropsychological test batteries. The primary aim of the present
investigation is to identify whether doxycycline negatively affects
declarative memory consolidation.
2.2. Participants
Participants were recruited from the general population and were
enrolled in one of the three RCTs (Bach et al., 2019; Wehrli et al., 2023;
Meister et al., under review). In brief, 261 healthy participants were
recruited from the general population. All participants were uent in
German and aged between 18 and 40 years. Experiments 1–2 included
participants of any biological sex; Experiment 3 female participants
only. Experiment 3 included only female participants to account for
gendered fear responses to the trauma lm paradigm measured as
intrusive recollections (Ferree et al., 2011). Participants were screened
for exclusion criteria such as medical/psychiatric disorders by a study
physician (for details see supplementary materials (SM) Table S1). Nine
participants were excluded from analysis after drug ingestion due to
vomiting, equipment failure, COVID-19 quarantine, drop-out, or
non-compliance with the study procedure (for details see Fig. 1). The
nal sample therefore included 252 participants: 126 in the placebo
group (mean age±SD: 24.38±4.50 years) and 126 in the doxycycline
group (mean age±SD: 24.27±4.07 years), see Table 1.
2.3. Trial registration
All studies were conducted in accordance with the Declaration of
Helsinki and approved by the governmental research ethics committee
(Kantonale Ethikkomission Zürich KEK-ZH-2014–0669 and KEK-ZH-
2018–01973) and the Swiss Agency for Therapeutic Products (Swiss-
medic, Bern, Switzerland; 2015DR1136 and 2019DR1026). All partici-
pants gave oral and written informed consent before the experiment
using forms approved by the ethics committee. Experiment 1 was pre-
registered at the primary ISRCTN registry (ISRCTN66987216) and at the
Swiss Federal Complementary Database (KOFAM; SNCTP000001439),
Experiment 2 and 3 were preregistered jointly at a WHO-approved
primary registry (German Clinical Trials Register, DRKS00017037)
and at the Swiss Federal Complementary Database (Kofam:
SNCTP000003485).
2.4. Study medication & randomization
The tetracyclic antibiotic doxycycline (brand name: Vibramycin®;
Pzer, Zurich Switzerland) was used as the study medication, and
mannitol as placebo. Doxycycline can cross the blood-brain-barrier and
is the detectable in the cerebrospinal uid (CSF) after 2–4 h (Dotevall
and Hagberg, 1989; Karlsson et al., 1996). A single oral dose of 200 mg
was used for all three experiments, corresponding to the smallest anti-
biotic dose recommended by the manufacturer and in line with a pre-
ceding study (Bach et al., 2018), which found reduction of fear memory
consolidation. The half-life of doxycycline is approximately 16 h, hence
by Visit 3 seven days after drug ingestion, the drug was washed out by
99.9%. The study medication was manufactured, blinded, and ran-
domized by a GMP-licensed pharmacy (Kantonsapotheke, Zurich,
Switzerland), separately for men and women. Study personnel were
blind to the randomization. Randomization for each study was broken
after the last participant completed the last visit.
2.5. Procedure
2.5.1. Visit 1: pre-experiment testing (day −14 - −1)
In the screening visit, a medical examination including blood and
urine tests was performed to check exclusion criteria (for details see SM
Table S1). In Experiment 1, an additional visit was conducted after Visit
1, seven days before drug administration (Visit 2) containing a
Pavlovian fear conditioning task. This additional visit did not entail any
neuropsychological tests.
2.5.2. Visit 2: neuropsychological testing under doxycycline (day 0)
In Visit 2, participants were again verbally screened for health issues
and then orally ingested the study medication. Experimental procedures
differed (for details see (Bach et al., 2019; Wehrli et al., 2023; Meister
et al., under review), however, the study drug was always administrated
3.5 to 4.5 h before neuropsychological tests started. Tests followed after
a brief fear memory reminder in experiment 1, after a 20-min fear
acquisition procedure in experiment 2, and directly after the metabo-
lization period in experiment 3. The neuropsychological testing pro-
cedure is detailed in Table 2.
2.5.3. Visit 3: delayed recall (day +7)
Seven days after Visit 2, participants returned for a prolonged
delayed recall visit, see Table 2.
J.M. Wehrli et al.
European Neuropsychopharmacology 89 (2024) 1–9
2

2.6. Cognitive test batteries
To assess declarative verbal memory, a German adaptation of the
Rey Auditory Verbal Learning Test the (RAVLT) was employed (version
A) (Helmstaedter et al., 2001; Lezak, 1976; Rey, 1941). In contrast to the
original RAVLT, an additional delayed recall (trial 7_day7) and delayed
recognition of list A (recognition_day7) after seven days on Visit 3 were
incorporated.
To assess declarative visual memory, the Rey Visual Design Learning
Task (RVDLT) (Rey, 1964) was used. To align the structure of the RVDLT
to the RAVLT a partial recreation of the German Figuraler Lern- und
Ged¨
achtnistest (B¨
aumler, 1974; see Balzer et al., 2011) was added to the
RVDLT, as an interference set B containing 15 new shapes. Analogous to
the RAVLT, a delayed recall (trial 7_day7) and recognition (recog-
nition_day7) of list A were repeated after seven days on Visit 3.
Procedural motor skill learning was assessed using the Finger
Sequence Tapping Task (FSTT) (Cellini, 2017). The learning phase
consists of 12 trials. After the learning phase, an immediate recall (3
trials) is performed. After a 20 min interval a delayed recall (3 trials) is
conducted. Equivalent to the declarative memory scores, a delayed
recall (3 trials) was carried out after seven days.
As a sustained attention test the continuous performance test (CPT)
(Knye et al., 2003), was conducted in Experiment 1 and 2. Additionally,
in Experiment 1 the digit symbol substitution test (DSST) was performed
to measure processing speed. Beyond processing speed, DSST perfor-
mance is inuenced by associative learning, visuo-perceptual function,
and working memory (Friedel, 2017; Jaeger, 2018).
2.7. Data analysis
Statistical analysis was performed in MATLAB (Version R2018a,
Math-Works) and Rstudio using R 4.0.2 (www.r-project.org, R Core
Assessed for
Eligibility (n = 334)
Excluded (n=73)
- not meeting study criteria (n=58)
- Declined participation (n=5)
- Other reasons (n=10)
Randomized
(n=261)
Allocated to
doxycycline (n=131)
Received doxycycline
(n=127)
Did not receive allocated
intervention:
- Vomiting (n=2)
- Technical issues (n=1)
- non-compliance (n=1)
Included in Analysis
(n=126)
Visit 3 missing:
- due to Covid-
Quarantine (n=1)
Allocated to placebo
(n=130)
Received placebo
(n=129)
Did not receive allocated
intervention:
- rescinded consent
(n=1)
Included in Analysis
(n=126)
Visit 3 missing:
- due to Covid-
Quarantine (n=2)
- technical issues in
Visit 3 (n=1)
Fig. 1. CONSORT Flow-Chart.
J.M. Wehrli et al.
European Neuropsychopharmacology 89 (2024) 1–9
3

Team, 2020).
2.8. Primary outcome
Combined test scores for RAVLT and RVDLT were computed as
general declarative memory scores. Based on the assumption that the
same molecular mechanisms of synaptic consolidation, which are
malleable to MMP9, underlie both verbal and visual declarative mem-
ory, a combined score could amplify the impact of doxycycline on
memory consolidation. Our main goal was to identify if doxycycline
would impede declarative memory retention as had been previously
stated in a study by Idzikowski and Oswald (1983). Because of the
varying study procedures, we included experiment as a factor. There-
fore, our primary analysis was a mixed repeated measures analyses of
variance (ANOVA) of the consolidation/retention phase (2 ×3*3 [dru-
g*experiments*trial 6, trial 7_day0, and trial 7_day7]). For the mixed
repeated measures ANOVA we used the function ezANOVA() of the R
package ‘ez’ 4.40.
2.9. Secondary outcomes
Because doxycycline was given before the learning phase we also
tested for effects on the learning (2 ×3*5 [drug*experiments*trials 1–5])
and recognition (2 ×3*2 [drug*experiments*recognition list A_day0 and
recognition list A_day7]) performance with a mixed repeated measures
ANOVA as secondary analyses. Furthermore, robustness analyses of
covariance (ANCOVA) accounting for age and sex were conducted for all
phases, i.e., consolidation, learning and recognition. Both for the
ANOVA and ANCOVA the function ezANOVA() of the R package ‘ez’
4.40 was employed. However, because ANCOVA analyses did not yield
any new insights, results are documented in the SM. Subsequently, if we
identied effects of drug in a phase, we further investigated group dif-
ferences by performing independent two sample t-tests comparing the
performance scores of the doxycycline group to the placebo group on the
measures. Calculations of performance scores for RAVLT and RVDLT are
detailed in SM Table S2. Finally, to identify test specic effects we also
conducted the analysis for the tests separately.
2.10. Control measures
For each FSTT phase (learning, immediate recall, delayed recall,
delayed recall after seven days) we obtained number of total and correct
sequences typed per trial. We conducted an mixed measures ANOVA and
ANCOVA including all phases using the function ezANOVA() of the R
package ‘ez’ 4.40: 2 ×21 (drug*12 learning trials, 9 recall trials [3 per
recall condition]). Because the ANCOVA analyses did not yield additional
insights they are documented in the SM. For group comparison (inde-
pendent t-test) a summarized score for each phase was calculated. Due to
technical issues, delayed recall after seven days was not available for
two participants in Experiment 1 and three participants in Experiment 3.
The outcome measures for the DSST were total and correct digits
substituted with a symbol, for the CPT hit and respective miss rate. For
both DSST and CPT we compared the outcomes between groups using
the Wilcoxon rank sum test as data failed the normality test (Shapiro-
Wilk parametric test). DSST data was only collected in Experiment 1 and
data of three participants was missing. CPT data was available for
Experiment 1 and 2, data of three participants was missing in Experi-
ment 1. No signicant differences were identied between placebo and
doxycycline group in the DSTT and CPT, see SM Table S13-S14, indi-
cating no impact on processing speed and sustained attention,
respectively.
3. Results
3.1. Primary outcome
Our primary outcome of interest was whether declarative memory
consolidation was signicantly reduced under doxycycline. In our pri-
mary ANOVA analysis of the consolidation phase, we nd no signicant
effect of the intervention. However, we nd a signicant effect of trials,
indicating a gain of remembered items between immediate recall to
delayed recall, followed by a subsequent loss of remembered items in the
delayed recall at Visit 3 seven days later. This analysis further indicated
signicant differences between experiments as participants remembered
signicantly less items in Experiment 3 compared to Experiment 1–2.
Further we identify an interaction effect for experiment and trials (see
Table 3).
3.2. Secondary outcomes
In our secondary analyses investigating whether doxycycline affects
the learning phase, we identify a better learning performance in the
doxycycline group in the rst few trials (Table 3 and Fig. 2). Addition-
ally, we identify an effect of trials (i.e., repeated exposure) showing an
increase of remembered items with each trial. Furthermore, we nd
Table 1
Overview of Participants age and sex distribution over the three experiments.
Experiment 1 (N =
78)
Placebo group (n =40) Doxycycline group (n =38)
male (n =
20)
female (n =
20)
male (n =
20)
female (n =
18)
mean age 25.00 ±
4.52
24.05 ±
5.82
25.44 ±
4.73
25.20 ±
5.30
Experiment 2 (N ¼
97)
Placebo group (n =48) Doxycycline group (n =49)
male (n =
24)
female (n =
24)
male (n =
24)
female (n =
25)
mean age 24.71 ±
3.57
25.08 ±
4.57
25.67 ±
4.00
23.24 ±
3.14
Experiment 3 (N ¼
77)
Placebo group (n =38) Doxycycline group (n =39)
male (n =
0)
female (n =
38)
male (n =
0)
female (n =
39)
mean age - 23.58 ±
4.23
- 23.05 ±
3.17
Total (N ¼252) Placebo group (n =126) Doxycycline group (n =126)
male (n =
44)
female (n =
82)
male (n =
44)
female (n =
82)
mean age 24.84 ±
3.94
24.13 ±
4.73
25.45 ±
4.53
23.63 ±
3.62
Table 2
Procedure for neuropsychological testing in experiment 1 to 3.
RAVLT: verbal learning and memory test; DSTT: Digit Symbol Substitution Test;
RVDLT: Rey visual Design Learning Task; FSTT: Finger Sequence Tapping Task;
CPT: Continuous Performance Test.
Visit 2 (day0)
Test Measures Experiments Time [minutes]
RAVLT; Part 1 Learning, interference trial &
immediate recall
1, 2, 3 10
DSTT Total & correct score 1 5
RVDLT; Part 1 Learning, interference trial &
immediate recall
1, 2, 3 17
FSTT; Part 1 Learning & immediate recall 1, 2, 3 15
CPT Hit/miss rate 1, 2 5
RAVLT; Part 2 Delayed recall and recognition 1, 2, 3 5
RVDLT; Part 2 Delayed recall and recognition 1, 2, 3 7
FSTT; Part 2 Delayed recall 1, 2, 3 3
Visit 3 (day7)
Test Measures Time [minutes]
RAVLT; Part 2 Delayed recall and recognition 1, 2, 3 6
DSST Total & correct score 1 5
RVDLT; Part 2 Delayed recall and recognition 1, 2, 3 7
FSTT; Part 2 Delayed recall 1, 2, 3 5
J.M. Wehrli et al.
European Neuropsychopharmacology 89 (2024) 1–9
4

interaction effects of drug and experiment as well as experiment and
trials.
In the recognition phase we nd a signicant effect of trials showing
that, after a prolonged absence of seven days, participants forget more of
the learned words. Further we nd an interaction of experiment and
trials, indicating different trajectories in the experiments (see Table 3).
In a direct comparison of declarative memory measures, we found
that the doxycycline group forgets less items after ~30 min consolida-
tion (loss after consolidation) than the placebo group (see Table 4).
Additionally, we found a signicant difference in the supraspan (trial 1),
whereby the doxycycline group remembered more items than the pla-
cebo group, for details see Table 4.
When conducting the ANOVA analysis for the RAVLT and RVDLT
separately for the consolidation phase we identify no main or interaction
effects of the drug. The effect of experiment as well as the interaction of
trial and experiment is only detectable in the RVDLT due to variance
between experiments, see Fig. 2c and SM Table S4-S7.
In the learning phase the main drug effect is only detectable in the
RVDLT but not RAVLT, conversely the interaction effect of trial and drug
in the learning phase indicating better learning performance in the
doxycycline group is only discernible in the RAVLT but not RVDLT, see
SM Table S4-S7. In the recognition phase we identify no main or inter-
action effects of drug in either test.
In a direct comparison of RAVLT measures we identify less loss after
Table 3
ANOVA analysis of general declarative memory (combined RAVLT & RVDLT).
Consolidation Learning Recognition
Effect df F-value p-value cohen’s d df F-value p-value cohen’s d df F-value p-value cohen’s d
Drug (doxycycline/placebo) 1, 245 1.04 0.31 0.07 1, 246 5.32 0.022* 0.15 1, 246 0.83 0.36 0.06
Experiment 2, 245 55.55 <0.001* 0.67 2, 246 82.35 <0.001* 0.82 2, 246 7.19 0.001* 0.24
Trials 2, 490 502.34 <0.001* 1.43 4, 984 1704.86 <0.001* 2.63 1, 246 136.46 <0.001* 0.74
Drug*Experiment 2, 245 2.11 0.12 0.13 2, 246 1.27 0.28 0.10 2, 246 1.47 0.23 0.11
Drug*Trial 2, 490 0.8 0.45 0.06 4, 984 3.76 0.005* 0.12 1, 246 0.08 0.78 0.02
Experiment*Trial 4, 490 11.48 <0.001* 0.31 8, 984 21.16 <0.001* 0.41 2, 246 4.54 0.012* 0.19
Drug*Experiment*Trial 4, 490 0.35 0.84 0.05 8, 984 1.4 0.19 0.11 2, 246 2.26 0.11 0.14
Fig. 2. a.) Combined RAVLT and RVDLT results, b.) for RAVLT and c.) for RVDLT Mean values of each group are depicted as lines ±SD in shaded areas. Mean values
per trial are additionally marked per experiment and per group as circles for experiment 1, squares for experiment 2 and triangles for experiment 3.
J.M. Wehrli et al.
European Neuropsychopharmacology 89 (2024) 1–9
5

consolidation on day 0 in the doxycycline group. For RVDLT no measure
signicantly differs between the doxycycline and placebo group, see SM
Table S10-S11.
3.3. Control measures
In the FSTT we identied an interaction effect of drug and trial both
for the total and the correct sequences over all phases (including
learning), indicating a smaller increase and decline in the doxycycline
group to and from the delayed recall on day 0 (see Fig. 3) meaning the
doxycycline group acquired the sequence less quickly but also retained
the sequence better in the delayed recall on day +7. We identify no
general drug effect. We nd no signicant differences in a direct
comparison of each FSTT session, see SM Table S12.
4. Discussion
The aim of this study was to assess whether a single dose of the
MMP9 inhibitor doxycycline negatively affects long-term declarative
memory consolidation in healthy human volunteers. We assessed visual
and verbal declarative memory in combination and separately. Addi-
tionally, non-declarative motor memory, sustained attention, and pro-
cessing were tested to check domain specicity of putative drug effects.
Over all experiments our results did not conrm a signicant
impairment of visual or verbal declarative memory in the doxycycline
group. Neither declarative memory consolidation nor declarative
Table 4
Mean data of general declarative memory (combined RAVLT & RVDLT) and t-test direct comparison of individual trial scores.
Placebo Doxycycline
Mean ±SD Mean ±SD df t-
value
p-value cohen’s d hedge’s g 95% CI
Supraspan (Trial A1_day0) 15.05 ±3.35 15.92 ±3.24 250 −2.10 0.037* −0.27 −0.27 [−1.69, −0.06]
Trial A2_day0 20.96 ±3.63 21.80 ±3.43 250 −1.89 0.060 −0.24 −0.24 [−1.72, 0.04]
Trial A3_day0 23.62 ±3.44 24.46 ±3.61 250 −1.89 0.059 −0.24 −0.24 [−1.72, 0.03]
Trial A4_day0 24.80 ±3.36 25.34 ±3.30 250 −1.29 0.20 −0.16 −0.16 [−1.37, 0.29]
Learning Performance (Trial A5_day0) 25.77 ±3.29 25.74 ±3.24 250 0.06 0.95 0.01 0.01 [−0.79, 0.83]
List B1_day0 16.15 ±4.08 16.38 ±3.96 250 −0.47 0.64 −0.06 −0.06 [−1.24, 0.76]
Trial A6_day0 24.45 ±3.45 24.50 ±3.71 250 −0.11 0.92 −0.01 −0.01 [−0.94, 0.84]
Trial A7_day0 24.84 ±3.65 25.25 ±3.64 250 −0.88 0.38 −0.11 −0.11 [−1.31, 0.50]
Trial A7_day7 19.75 ±4.52 20.21 ±4.69 249 −0.78 0.44 −0.10 −0.10 [−1.60, 0.69]
Total Learning Performance (∑Trial A1–5_day0) 110.19 ±15.03 113.26 ±15.04 250 −1.62 0.11 −0.21 −0.20 [−6.80, 0.66]
Proactive Interference (Trial A1_day0 - trial B1_day0) −1.10 ±3.06 −0.46 ±2.89 250 −1.69 0.091 −0.21 −0.21 [−1.37, 0.10]
Retroactive Interference (Trial A5_day0 - A6_day0) 1.32 ±1.69 1.25 ±1.91 250 0.31 0.75 0.04 0.04 [−0.38, 0.52]
Loss day0 (Trial A5_day0 -A7_day0) 0.92 ±1.66 0.50 ±1.63 250 2.07 0.040* 0.26 0.26 [0.02, 0.84]
Loss day7 (Trial A5_day0 -A7_day7) 6.04 ±3.70 5.54 ±3.67 249 1.09 0.28 0.14 0.14 [−0.41, 1.42]
Recognition day0 28.90 ±1.54 29.09 ±1.11 250 −1.08 0.28 −0.14 −0.14 [−0.52, 0.15]
Recognition day7 27.63 ±2.06 27.75 ±1.92 250 −0.44 0.66 −0.06 −0.06 [−0.61, 0.38]
Fig. 3. Recall results (averaged over phases) of the FSTT for all experiments.
J.M. Wehrli et al.
European Neuropsychopharmacology 89 (2024) 1–9
6

learning were negatively affected by doxycycline in any modality.
Contrarily, the doxycycline group shows improved declarative learning
and less loss after consolidation on the day of drug ingestion (Visit 2). An
interaction effect of drug and trials indicates that the doxycycline group
remembered more items specically in the early learning phase. While
the learning improvement was mainly driven by the visual memory test,
the reduced loss after consolidation was mostly driven by the verbal
memory test.
The results for the motor memory task identied a signicant
interaction effect of drug and trials over all phases. While the placebo
and doxycycline groups show similar levels during the learning phase
and at the delayed recall on day +7, the placebo group shows an in-
crease at the delayed recall on day 0, which is less pronounced in the
doxycycline group. However, doxycycline did not impact sustained
attention (measured with the CPT) and processing speed (measured with
the DSST) in any of the experiments or combined analyses. Of note, the
DSST is particularly sensitive to drug challenges and covers also higher
cognitive processing (Jaeger, 2018); thus, our ndings seem to reect a
specic memory effect of doxycycline. It is safe to assume that our study
sample was representative of the tested population; all averaged test
scores of the placebo group where within the normal range of the
assessed tests (Balzer et al., 2011; Helmstaedter et al., 2001; Petermann
and Wechsler, 2014).
Our results concerning declarative measure differ from the ndings
of the study by Idzikowskil and Oswald (1983), which demonstrated a
decrease in verbal memory capability after doxycycline consumption.
However, the study by Idzikowskil and Oswald tested verbal declarative
memory differently using 40-word sentences, which had to be learned
and remembered whilst REM sleep was manipulated. Furthermore, the
sample consisted of only 32 participants (16 per group; mean age 21.60
years), which is much smaller compared to our total cohort of 252
participants (126 per group; mean age 24.38 years).
Our hypothesis was based on the assumption that MMP9 inhibition
would show a detrimental impact on declarative memory as MMP9 in-
hibition has previously been shown to interfere with fear conditioning
reconsolidation in rodents (Brown et al., 2009) and fear memory
consolidation in humans (Bach et al., 2018). Additionally, MMP9 inhi-
bition has been shown to negatively impact hippocampus-dependent
spatial learning in animals (Wright et al., 2007), though, this has so
far not been conrmed in humans. We identify no attenuating effect on
declarative memory in our results. One possible explanation for our
ndings is that MMP9 might not be as critical for human
hippocampus-dependent memory consolidation as we assumed.
Furthermore, doxycycline inuences also other MMPs than MMP9
(Burggraf et al., 2007; Golub et al., 1991), which may obfuscate the
effect pure MMP9 inhibition may have on declarative memory.
Another cellular mechanism underlying learning and memory
affected by doxycycline is inhibition of microglial activation. Microglia
regulate synaptic plasticity and are therefore assumed to be critically
involved in learning and memory (Wang et al., 2020). Abnormal
microglial activation and potential subsequent neuroinammation are
crucial mechanisms involved with cognitive decline associated with
aging or neurogenerative diseases (Cornell et al., 2022). In Alzheimer’s
disease mouse models inhibition of microglia is assumed to halt tau
secretion and decelerate neurodegeneration (Gao et al., 2023; Sugama
et al., 2009), further amyloid-beta is suggested to induce
microglia-mediated neuroinammation (Qiao et al., 2021). Microglia
eliminate surplus synapses in the developing brain; therefore, it is
assumed that the elimination of synaptic connections in the adult hip-
pocampus may be one of the key mechanisms underlying forgetting.
Consequently, microglial inhibition has also been tested as a potential
mechanism to prevent forgetting (Klein, 2020). Experiments in rodents
have shown that depletion or inhibition of microglial activation im-
proves spatial and fear memory, (Gentry et al., 2022; Wadhwa et al.,
2017; Wang et al., 2020).
Our ndings are convergent with the assumption that microglial
inhibition would support learning and memory. Although we only
detected small effect sizes for the doxycycline effect, our results suggest
that doxycycline subtly but robustly increases the performance in
declarative memory tasks. In the reverse case, it is safe to assume that a
single oral dose of 200-mg doxycycline is not enough to affect declara-
tive memory negatively. Notably, doxycycline targets many cellular
mechanisms involved in the learning and consolidation of memory,
therefore, it is not possible to disentangle the effects with the study at
hand. We propose that declarative memory-enhancing of doxycycline
could be explained by microglial inhibition (Gao et al., 2023), which has
to be validated in future studies. Future studies investigating the effects
of doxycycline on memory consolidation should consider recruiting
older adults. Our study employed young adults, however older adults
have higher microglia activation (Conde and Streit, 2006; Norden and
Godbout, 2013), which may lead to stronger effects of tetracyclines on
memory. Finally, testing for effects of doxycycline on declarative
memory during prolonged medication intake – according to neuro-
protective treatment schemes – should be considered as a next step.
Given that microglial mechanisms are currently discussed in the treat-
ment of several neurodegenerative diseases, the clinical relevance of
such studies is self-evident.
Contributors
BBQ, DRB & BK concepted the studies. BBQ, DRB, YX, LM & JW
designed the studies. YX, LM, ST & JW acquired the data. ST & JW
managed the literature searches. BBQ, DRB & JW undertook the statis-
tical analysis and JW wrote the rst draft of the manuscript. BBQ & DRB
revised the manuscript. All authors contributed to and have approved
the nal manuscript.
Role of funding source
This work was supported by the Clinical Research Priority Program
of the University of Zurich for the CRPP "Synapse & Trauma" and his
successor “Synapse, Trauma, and Addiction”. The Sponsor of Studies 2 &
3 was the Department of Psychiatry, Psychotherapy, Psychosomatics,
University of Zurich (Prof. Erich Seifritz), the sponsor of Study 1 was
Prof. Dominik Bach. Prof. Seifritz had no further role in study design; in
the collection, analysis and interpretation of data; in the writing of the
report; and in the decision to submit the paper for publication. Prof.
Bach was involved in study design; in the collection, analysis and
interpretation of data; in the writing of the report; and in the decision to
submit the paper for publication.
Declaration of competing interest
The authors declare no competing nancial interests.
Acknowledgments
We thank all Bachelor and Master students associated with the ex-
periments for their help in data acquisition and Samuel Gerster for
technical support. This work was supported by the Clinical Research
Priority Program of the University of Zurich for the CRPP "Synapse &
Trauma" and his successor “Synapse, Trauma, and Addiction”.
Supplementary materials
Supplementary material associated with this article can be found, in
the online version, at doi:10.1016/j.euroneuro.2024.08.006.
References
Albouy, G., Fogel, S., King, B.R., Laventure, S., Benali, H., Karni, A., Carrier, J.,
Robertson, E.M., Doyon, J., 2015. Maintaining vs. enhancing motor sequence
J.M. Wehrli et al.
European Neuropsychopharmacology 89 (2024) 1–9
7

memories: respective roles of striatal and hippocampal systems. Neuroimage 108,
423–434. https://doi.org/10.1016/j.neuroimage.2014.12.049.
Bach, D.R., N¨
af, M., Deutschmann, M., Tyagarajan, S.K., Quednow, B.B., 2019. Threat
memory reminder under matrix metalloproteinase 9 inhibitor doxycycline globally
reduces subsequent memory plasticity. J. Neurosci. 39 (47), 9424–9434. https://doi.
org/10.1523/JNEUROSCI.1285-19.2019.
Bach, D.R., Tzovara, A., Vunder, J., 2018. Blocking human fear memory with the matrix
metalloproteinase inhibitor doxycycline. Mol. Psychiatry 23 (7), 1584–1589.
https://doi.org/10.1038/mp.2017.65.
Balducci, C., Forloni, G., 2019. Doxycycline for Alzheimer’s Disease: ghting β-Amyloid
Oligomers and Neuroinammation. Front. Pharmacol. 10, 738. https://doi.org/
10.3389/fphar.2019.00738.
Balducci, C., Santamaria, G., La Vitola, P., Brandi, E., Grandi, F., Viscomi, A.R., Beeg, M.,
Gobbi, M., Salmona, M., Ottonello, S., Forloni, G., 2018. Doxycycline counteracts
neuroinammation restoring memory in Alzheimer’s disease mouse models.
Neurobiol. Aging 70, 128–139. https://doi.org/10.1016/j.
neurobiolaging.2018.06.002.
Balzer, C., Berger, J.-M., Caprez, G., Gonsor, A., Gutbrod, K., Keller, M., 2011. MNND -
Materialien und Normwerte für die Neuropsychologische Diagnostik. Verlag,
Normdaten.
B¨
aumler, G., 1974. Lern- und Ged¨
achtnistest. Verlag für Psychologie.
Berens, S.C., Bird, C.M., Harrison, N.A., 2020. Minocycline differentially modulates
human spatial memory systems. Neuropsychopharmacology 45 (13), 2162–2169.
https://doi.org/10.1038/s41386-020-00811-8.
Bernardino, A.L.F., Kaushal, D., Philipp, M.T., 2009. The antibiotics doxycycline and
minocycline inhibit the inammatory responses to the lyme disease spirochete
Borrelia burgdorferi. J. Infect. Dis. 199 (9), 1379–1388. https://doi.org/10.1086/
597807.
Bortolanza, M., Nascimento, G.C., Socias, S.B., Ploper, D., Chehín, R.N., Raisman-
Vozari, R., Del-Bel, E., 2018. Tetracycline repurposing in neurodegeneration: focus
on Parkinson’s disease. J. Neural Transm. 125 (10), 1403–1415. https://doi.org/
10.1007/s00702-018-1913-1.
Boyle, J.R., McDermott, E., Crowther, M., Wills, A.D., Bell, P.R.F., Thompson, M.M.,
1998. Doxycycline inhibits elastin degradation and reduces metalloproteinase
activity in a model of aneurysmal disease. J. Vasc. Surg. 27 (2), 354–361. https://
doi.org/10.1016/S0741-5214(98)70367-2.
Brown, T.E., Wilson, A.R., Cocking, D.L., Sorg, B.A., 2009. Inhibition of matrix
metalloproteinase activity disrupts reconsolidation but not consolidation of a fear
memory. Neurobiol. Learn. Mem. 91 (1), 66–72. https://doi.org/10.1016/j.
nlm.2008.09.003.
Burggraf, D., Trinkl, A., Dichgans, M., Hamann, G.F., 2007. Doxycycline inhibits MMPs
via modulation of plasminogen activators in focal cerebral ischemia. Neurobiol. Dis.
25 (3), 506–513. https://doi.org/10.1016/j.nbd.2006.10.013.
Cellini, N., 2017. Memory consolidation in sleep disorders. Sleep. Med. Rev. 35,
101–112. https://doi.org/10.1016/j.smrv.2016.09.003.
Conde, J.R., Streit, W.J., 2006. Microglia in the aging brain. J. Neuropathol. Exp. Neurol.
65 (3), 199–203. https://doi.org/10.1097/01.jnen.0000202887.22082.63.
Cornell, J., Salinas, S., Huang, H.-Y., Zhou, M., 2022. Microglia regulation of synaptic
plasticity and learning and memory. Neural Regen. Res. 17 (4), 705. https://doi.org/
10.4103/1673-5374.322423.
Dotevall, L., Alestig, K., Hanner, P., Norkrans, G., Hagberg, L., 1988. The use of
doxycycline in nervous system Borrelia burgdorferi infection. Scand. J. Infect. Dis.
Suppl. 53, 74–79.
Dotevall, L., Hagberg, L., 1989. Penetration of doxycycline into cerebrospinal uid in
patients treated for suspected Lyme neuroborreliosis. Antimicrob. Agents
Chemother. 33 (7), 1078–1080. https://doi.org/10.1128/AAC.33.7.1078.
Eichenbaum, H., 2004. Hippocampus. Neuron 44 (1), 109–120. https://doi.org/
10.1016/j.neuron.2004.08.028.
Ferree, N.K., Kamat, R., Cahill, L., 2011. Inuences of menstrual cycle position and sex
hormone levels on spontaneous intrusive recollections following emotional stimuli.
Conscious. Cogn. 20 (4), 1154–1162. https://doi.org/10.1016/j.
concog.2011.02.003.
Fischer, J., Ganellin, C.R (Eds.), 2006. Analogue-Based Drug Discovery. Wiley-VCH.
Friedel, E., 2017. How Accumulated Real Life Stress Experience and Cognitive Speed
Interact on Decision-Making Processes. Front. Hum. Neurosci. 11 https://doi.org/
10.3389/fnhum.2017.00302.
Gao, C., Jiang, J., Tan, Y., Chen, S., 2023. Microglia in neurodegenerative diseases:
mechanism and potential therapeutic targets. Signal. Transduct. Target. Ther. 8 (1),
359. https://doi.org/10.1038/s41392-023-01588-0.
Gentry, N.W., McMahon, T., Yamazaki, M., Webb, J., Arnold, T.D., Rosi, S., Pt´
aˇ
cek, L.J.,
Fu, Y.-H., 2022. Microglia are involved in the protection of memories formed during
sleep deprivation. Neurobiol. Sleep. Circadian. Rhythms. 12, 100073 https://doi.
org/10.1016/j.nbscr.2021.100073.
Gheysen, F., Van Opstal, F., Roggeman, C., Van Waelvelde, H., Fias, W., 2010.
Hippocampal contribution to early and later stages of implicit motor sequence
learning. Exp. Brain Res. 202 (4), 795–807. https://doi.org/10.1007/s00221-010-
2186-6.
Golub, L.M., Ramamurthy, N.S., McNamara, T.F., Greenwald, R.A., Rifkin, B.R., 1991.
Tetracyclines inhibit connective tissue breakdown: new therapeutic implications for
an old family of drugs. Crit. Rev. Oral Biol. Med. 2 (3), 297–321. https://doi.org/
10.1177/10454411910020030201.
Gould, T.J., Rowe, W.B., Heman, K.L., Mesches, M.H., Young, D.A., Rose, G.M.,
Bickford, P.C., 2002. Effects of hippocampal lesions on patterned motor learning in
the rat. Brain Res. Bull. 58 (6), 581–586. https://doi.org/10.1016/S0361-9230(02)
00832-8.
Greenblatt, D.J., Gan, L., Harmatz, J.S., Shader, R.I., 2005. Pharmocokinetics and
pharmacodynamics of single-dose triazolam: electroencephalography compared
with the Digit-Symbol Substitution Test. Br. J. Clin. Pharmacol. 60 (3), 244–248.
https://doi.org/10.1111/j.1365-2125.2005.02409.x.
Gross, K.S., Lincoln, C.M., Anderson, M.M., Geiger, G.E., Frick, K.M., 2022. Extracellular
matrix metalloproteinase-9 (MMP-9) is required in female mice for 17β-estradiol
enhancement of hippocampal memory consolidation. Psychoneuroendocrinology.
141, 105773 https://doi.org/10.1016/j.psyneuen.2022.105773.
Helmstaedter, C., Lendt, M., Lux, S., 2001. Verbaler Lern- und Merkf¨
ahigkeitstest,
Manual. Beltz.
Idzikowski, C., Oswald, I., 1983. Interference with human memory by an antibiotic.
Psychopharmacology (Berl.) 79 (2–3), 108–110. https://doi.org/10.1007/
BF00427794.
Jacobacci, F., Armony, J.L., Yeffal, A., Lerner, G., Amaro, E., Jovicich, J., Doyon, J.,
Della-Maggiore, V., 2020. Rapid hippocampal plasticity supports motor sequence
learning. In: Proceedings of the National Academy of Sciences, 117,
pp. 23898–23903. https://doi.org/10.1073/pnas.2009576117.
Jaeger, J., 2018. Digit symbol substitution test: the case for sensitivity over specicity in
neuropsychological testing. J. Clin. Psychopharmacol. 38 (5), 513–519. https://doi.
org/10.1097/JCP.0000000000000941.
James, E.L., Lau-Zhu, A., Clark, I.A., Visser, R.M., Hagenaars, M.A., Holmes, E.A., 2016.
The trauma lm paradigm as an experimental psychopathology model of
psychological trauma: intrusive memories and beyond. Clin. Psychol. Rev. 47,
106–142. https://doi.org/10.1016/j.cpr.2016.04.010.
Karlsson, M., Hammers, S., Nilsson-Ehle, I., Malmborg, A.S., Wretlind, B., 1996.
Concentrations of doxycycline and penicillin G in sera and cerebrospinal uid of
patients treated for neuroborreliosis. Antimicrob. Agents Chemother. 40 (5),
1104–1107. https://doi.org/10.1128/AAC.40.5.1104.
Kim, H.-S., Luo, L., Pugfelder, S.C., Li, D.-Q., 2005. Doxycycline inhibits TGF-
β1–induced MMP-9 via Smad and MAPK pathways in human corneal epithelial cells.
Investig. Opthalmol. Vis. Sci. 46 (3), 840. https://doi.org/10.1167/iovs.04-0929.
Klein, R.S., 2020. On complement, memory, and microglia. N. Engl. J. Med. 382 (21),
2056–2058. https://doi.org/10.1056/NEJMcibr2002480.
Knye, M., Roth, N., Westhus, W., Heine, A., 2003. Continuous performance test. Hogrefe.
Lezak, M.D., 1976. Neuropsychological Assessment. Oxford University Press.
Miyachi, Y., Yoshioka, A., Imamura, S., Niwa, Y., 1986. Effect of antibiotics on the
generation of reactive oxygen species. J. Investig. Dermatol. 86 (4), 449–453.
https://doi.org/10.1111/1523-1747.ep12285793.
Nagy, V., 2006. Matrix metalloproteinase-9 is required for hippocampal late-phase long-
term potentiation and memory. J. Neurosci. 26 (7), 1923–1934. https://doi.org/
10.1523/JNEUROSCI.4359-05.2006.
Nagy, V., Bozdagi, O., Huntley, G.W., 2007. The extracellular protease matrix
metalloproteinase-9 is activated by inhibitory avoidance learning and required for
long-term memory. Learn. Mem. 14 (10), 655–664. https://doi.org/10.1101/
lm.678307.
Nelson, M.L., Levy, S.B., 2011. The history of the tetracyclines: the history of the
tetracyclines. Ann. N. Y. Acad. Sci. 1241 (1), 17–32. https://doi.org/10.1111/
j.1749-6632.2011.06354.x.
Norden, D.M., Godbout, J.P., 2013. Review: microglia of the aged brain: primed to be
activated and resistant to regulation. Neuropathol. Appl. Neurobiol. 39 (1), 19–34.
https://doi.org/10.1111/j.1365-2990.2012.01306.x.
Ogg, R.J., Zou, P., Allen, D.N., Hutchins, S.B., Dutkiewicz, R.M., Mulhern, R.K., 2008.
Neural correlates of a clinical continuous performance test. Magn. Reson. ImAging
26 (4), 504–512. https://doi.org/10.1016/j.mri.2007.09.004.
Orsucci, D., Calsolaro, V., Mancuso, M., Siciliano, G., 2009. Neuroprotective effects of
tetracyclines: molecular targets. Anim. Models Hum. Dis. CNS Neurol. Disord. - Drug
Targets 8 (3), 222–231. https://doi.org/10.2174/187152709788680689.
Petermann, F., Wechsler, D. (Eds.), 2014. Wechsler Adult Intelligence Scale (WAIS-IV).
Manual 1: Grundlagen, Testauswertung und Interpretation /Übersetzung und
Adaptation der WAIS-IV Von David Wechsler ; F. Petermann (Hrsg.) (2., Erweiterte
Auage). Pearson.
Qiao, O., Ji, H., Zhang, Y., Zhang, X., Zhang, X., Liu, N., Huang, L., Liu, C., Gao, W., 2021.
New insights in drug development for Alzheimer’s disease based on microglia
function. Biomed. Pharmacother. 140, 111703 https://doi.org/10.1016/j.
biopha.2021.111703.
Rahmani, M., Negro ´
Alvarez, S.E., Hern´
andez, E.B., 2022. The potential use of
tetracyclines in neurodegenerative diseases and the role of nano-based drug delivery
systems. Eur. J. Pharmaceut. Sci. 175, 106237 https://doi.org/10.1016/j.
ejps.2022.106237.
Ramamurthy, N.S., Vernillo, A.T., Greenwald, R.A., Lee, H.-M., Sorsa, T., Golub, L.M.,
Rifkin, B.R., 2009. Reactive oxygen species activate and tetracyclines inhibit rat
osteoblast collagenase. J. Bone Miner. Res. 8 (10), 1247–1253. https://doi.org/
10.1002/jbmr.5650081013.
Rey, A. (1941). L’examen psychologique dans les cas d’encephopathie traumatique (The
psychological examination of cases of traumatic encephalopathy). 28, 286–340.
Rey, A., 1964. L’examen Clinique Enpsychologie. Presses Universitaires de France.
Riccio, C.A., Reynolds, C.R., Lowe, P., Moore, J.J., 2002. The continuous performance
test: a window on the neural substrates for attention? Arch. Clin. Neuropsychol. 17
(3), 235–272. https://doi.org/10.1093/arclin/17.3.235.
Santa-Cecília, F.V., Leite, C.A., Del-Bel, E., Raisman-Vozari, R., 2019. The
neuroprotective effect of doxycycline on neurodegenerative diseases. Neurotox. Res.
35 (4), 981–986. https://doi.org/10.1007/s12640-019-00015-z.
Santa-Cecília, F.V., Socias, B., Ouidja, M.O., Sepulveda-Diaz, J.E., Acu˜
na, L., Silva, R.L.,
Michel, P.P., Del-Bel, E., Cunha, T.M., Raisman-Vozari, R., 2016. Doxycycline
suppresses microglial activation by inhibiting the p38 MAPK and NF-kB signaling
J.M. Wehrli et al.
European Neuropsychopharmacology 89 (2024) 1–9
8

pathways. Neurotox. Res. 29 (4), 447–459. https://doi.org/10.1007/s12640-015-
9592-2.
Sapadin, A.N., Fleischmajer, R., 2006. Tetracyclines: nonantibiotic properties and their
clinical implications. J. Am. Acad. Dermatol. 54 (2), 258–265. https://doi.org/
10.1016/j.jaad.2005.10.004.
Schapiro, A.C., Reid, A.G., Morgan, A., Manoach, D.S., Verfaellie, M., Stickgold, R., 2019.
The hippocampus is necessary for the consolidation of a task that does not require
the hippocampus for initial learning. Hippocampus 29 (11), 1091–1100. https://doi.
org/10.1002/hipo.23101.
Sugama, S., Takenouchi, T., Cho, B., Joh, T., Hashimoto, M., Kitani, H., 2009. Possible
roles of microglial cells for neurotoxicity in clinical neurodegenerative diseases and
experimental animal models. Inamm. Allergy-Drug Targets 8 (4), 277–284. https://
doi.org/10.2174/187152809789352249.
Thapar, P., Zacny, J.P., Thompson, W., Apfelbaum, J.L., 1995. Using alcohol as a
standard to assess the degree of impairment induced by sedative and analgesic drugs
used in ambulatory surgery. Anesthesiology 82 (1), 53–59. https://doi.org/10.1097/
00000542-199501000-00008.
Vafadari, B., Salamian, A., Kaczmarek, L., 2016. MMP-9 in translation: from molecule to
brain physiology, pathology, and therapy. J. Neurochem. 139, 91–114. https://doi.
org/10.1111/jnc.13415.
Wadhwa, M., Prabhakar, A., Ray, K., Roy, K., Kumari, P., Jha, P.K., Kishore, K.,
Kumar, S., Panjwani, U., 2017. Inhibiting the microglia activation improves the
spatial memory and adult neurogenesis in rat hippocampus during 48 h of sleep
deprivation. J. Neuroinammation. 14 (1), 222. https://doi.org/10.1186/s12974-
017-0998-z.
Wang, C., Yue, H., Hu, Z., Shen, Y., Ma, J., Li, J., Wang, X.-D., Wang, L., Sun, B., Shi, P.,
Wang, L., Gu, Y., 2020. Microglia mediate forgetting via complement-dependent
synaptic elimination. Science (1979) 367 (6478), 688–694. https://doi.org/
10.1126/science.aaz2288.
Wehrli, J.M., Xia, Y., Offenhammer, B., Kleim, B., Müller, D., Bach, D.R., 2023. Effect of
the matrix metalloproteinase inhibitor doxycycline on human trace fear memory.
eNeuro 10 (2). https://doi.org/10.1523/ENEURO.0243-22.2023. ENEURO.0243-
22.2023.
Wright, J.W., Brown, T.E., Harding, J.W., 2007. Inhibition of hippocampal matrix
metalloproteinase-3 and -9 disrupts spatial memory. Neural Plast. 2007, 1–8.
https://doi.org/10.1155/2007/73813.
Xia, Y., Wehrli, J., Abivardi, A., Hostiuc, M., Kleim, B., Bach, D.R., 2024. Attenuating
human fear memory retention with minocycline: a randomized placebo-controlled
trial. Transl. Psychiatry 14 (1), 28. https://doi.org/10.1038/s41398-024-02732-2.
Yrj¨
anheikki, J., Kein¨
anen, R., Pellikka, M., H¨
okfelt, T., Koistinaho, J., 1998. Tetracyclines
inhibit microglial activation and are neuroprotective in global brain ischemia. Proc.
Natl. Acad. Sci. 95 (26), 15769–15774. https://doi.org/10.1073/pnas.95.26.15769.
J.M. Wehrli et al.
European Neuropsychopharmacology 89 (2024) 1–9
9

Forget me not: The effect of doxycycline on human declarative memory 1
Table S1
Participant Selection
Inclusion criteria
All Studies
Healthy, German-speaking, between 18 and 40 years of age
Only Experiment 3
Female Sex
Exclusion criteria
Contraindication to doxycycline or mannitol (history of allergic reactions)
Drug use in the two weeks prior to the study (exception: contraceptive drugs and incidental use of NSARs/paracetamol)
Known or suspected use of illicit drugs, use of benzodiazepines, alcohol abuse
Any history of psychiatric, neurological, or systemic/rheumatic disease
Other clinically significant concomitant disease states
Pregnant or breast-feeding women, intention to become pregnant, lack of safe contraception method
Participation in any other drug study within the 30 days preceding and during the present study
Blood parameters screened
Blood cell count, electrolytes
C-reactive protein
Aspartate aminotransferase (ASAT/GOT)
Alanine aminotransferase (ALAT/GPT)
Gamma-glutamyl transferase (gamma-GT)
Kreatinin,
Thyroid-stimulating hormone (TSH),
Free thyroxine (FT4)
Urine parameters screened
Amphetamines
Barbiturates
Benzodiazepines
Tetrahydrocannabinol
Cocaine
Methadone
Opioids
Additionally only for women
beta human chorionic gonadotropin (beta-HCG) pregnancy test

Forget me not: The effect of doxycycline on human declarative memory 2
Table S2
Performance Scores RAVLT & RVDLT
Measured Variable Variables
Supraspan Trial 1
Learning Performance Trial 5
Total Learning Performance ∑Trial 1-5
Immediate Recall Trial 6
Delayed Recall (30min) Trial 7_day0
Delayed Recall (7days) Trial 7_day7
Loss after consolidation (30min) Trial 5 - 7_day0
Loss after consolidation (7days) Trial 5 - 7_day7
Interference Performance Trial B
Pro-active Interference Trial 1 - B
Retro-active Interference Trial 5 - 6
Recognition after consolidation (30min) Recognition_day0
Recognition after consolidation (7days) Recognition_day7

Forget me not: The effect of doxycycline on human declarative memory 3
Table S3
ANCOVA RAVLT+RVDLT Combined with Experiment as between subject Factor
Consolidation Recognition
Effect F-value p-value cohen's d F-value p-value cohen's d F-value p-value cohen's d
Drug (doxycycline/placebo) 1 , 245 1.06 0.30 0.07 1 , 246 5.59 0.019* 0.15 1 , 246 0.84 0.36 0.06
Experiment 2 , 245 39.31 <.001* 0.57 2 , 246 65.23 <.001* 0.73 2 , 246 6.67 0.002* 0.23
Trials 2 , 490 502.34 <.001* 1.43 4 , 984 1704.86 <.001* 2.63 1 , 246 136.46 <.001* 0.74
Drug x Experiment 2 , 245 2.07 0.13 0.13 2 , 246 1.55 0.21 0.11 2 , 246 1.51 0.22 0.11
Drug x Trial 2 , 490 0.80 0.45 0.06 4 , 984 3.76 0.004* 0.12 1 , 246 0.08 0.78 0.02
Experiment x Trial 4 , 490 11.48 <.001* 0.31 8 , 984 21.16 <.001* 0.41 2 , 246 4.54 0.012* 0.19
Drug x Experiment x Trial 4 , 490 0.35 0.84 0.05 8 , 984 1.40 0.19 0.11 2 , 246 2.26 0.11 0.14
Table S4
ANOVA RAVLT
Consolidation Recognition
Effect F-value p-value cohen's d F-value p-value cohen's d F-value p-value cohen's d
Drug (doxycycline/placebo) 1 , 246 0.22 0.64 0.03 1 , 246 1.14 0.29 0.07 1 , 246 0.93 0.34 0.06
Experiment 2 , 246 0.43 0.65 0.06 2 , 246 0.99 0.37 0.09 2 , 246 3.49 0.032* 0.17
Trials 2 , 492 311.83 <.001* 1.12 4 , 984 724.69 <.001* 1.71 1 , 246 80.60 <.001* 0.57
Drug x Experiment 2 , 246 1.02 0.36 0.09 2 , 246 0.88 0.42 0.08 2 , 246 0.32 0.73 0.05
Drug x Trial 2 , 492 1.36 0.26 0.07 4 , 984 3.26 0.011* 0.11 1 , 246 0.11 0.74 0.02
Experiment x Trial 4 , 492 0.73 0.57 0.08 8 , 984 0.85 0.55 0.08 2 , 246 3.55 0.030* 0.17
Drug x Experiment x Trial 4 , 492 0.22 0.93 0.04 8 , 984 1.10 0.36 0.09 2 , 246 0.81 0.44 0.08
df
df
df
df
df
df
Learning
Learning

Forget me not: The effect of doxycycline on human declarative memory 4
Table S5
ANCOVA RAVLT
Consolidation Recognition
Effect F-value p-value cohen's d F-value p-value cohen's d F-value p-value cohen's d
Drug (doxycycline/placebo) 1 , 246 0.22 0.64 0.03 1 , 246 1.16 0.28 0.07 1 , 246 0.95 0.33 0.06
Experiment 2 , 246 0.53 0.59 0.07 2 , 246 1.61 0.20 0.11 2 , 246 2.22 0.11 0.13
Trials 2 , 492 311.83 <.001* 1.12 4 , 984 724.69 <.001* 1.71 1 , 246 80.60 <.001* 0.57
Drug x Experiment 2 , 246 1.03 0.36 0.09 2 , 246 0.96 0.38 0.09 2 , 246 0.43 0.65 0.06
Drug x Trial 2 , 492 1.36 0.26 0.07 4 , 984 3.26 0.011* 0.11 1 , 246 0.11 0.74 0.02
Experiment x Trial 4 , 492 0.73 0.57 0.08 8 , 984 0.85 0.55 0.08 2 , 246 3.55 0.030* 0.17
Drug x Experiment x Trial 4 , 492 0.22 0.93 0.04 8 , 984 1.10 0.36 0.09 2 , 246 0.81 0.44 0.08
Table S6
ANOVA RVDLT
Consolidation Recognition
Effect F-value p-value cohen's d F-value p-value cohen's d F-value p-value cohen's d
Drug (doxycycline/placebo) 1 , 245 1.76 0.19 0.08 1 , 246 6.89 0.009* 0.17 1 , 246 0.17 0.68 0.03
Experiment 2 , 245 151.40 <.001* 1.11 2 , 246 179.97 <.001* 1.21 2 , 246 5.46 0.005* 0.21
Trials 2 , 490 302.07 <.001* 1.11 4 , 984 1107.49 <.001* 2.12 1 , 246 72.68 <.001* 0.54
Drug x Experiment 2 , 245 1.93 0.15 0.13 2 , 246 0.81 0.45 0.08 2 , 246 2.58 0.078 0.14
Drug x Trial 2 , 490 0.25 0.78 0.03 4 , 984 1.64 0.16 0.08 1 , 246 0.74 0.39 0.05
Experiment x Trial 4 , 490 21.23 <.001* 0.42 8 , 984 36.83 <.001* 0.55 2 , 246 1.67 0.19 0.12
Drug x Experiment x Trial 4 , 490 0.36 0.84 0.05 8 , 984 1.59 0.12 0.11 2 , 246 1.96 0.14 0.13
df
df
df
df
df
df
Learning
Learning

Forget me not: The effect of doxycycline on human declarative memory 5
Table S7
ANCOVA RVDLT
Consolidation Recognition
Effect F-value p-value cohen's d F-value p-value cohen's d F-value p-value cohen's d
Drug (doxycycline/placebo) 1 , 245 1.68 0.20 0.08 1 , 246 6.83 0.010* 0.17 1 , 246 0.16 0.69 0.03
Experiment 2 , 245 106.03 <.001* 0.93 2 , 246 135.88 <.001* 1.05 2 , 246 5.77 0.004* 0.22
Trials 2 , 490 302.07 <.001* 1.11 4 , 984 1107.49 <.001* 2.12 1 , 246 72.68 <.001* 0.54
Drug x Experiment 2 , 245 1.68 0.19 0.12 2 , 246 1.03 0.36 0.09 2 , 246 2.46 0.088 0.14
Drug x Trial 2 , 490 0.25 0.78 0.03 4 , 984 1.64 0.16 0.08 1 , 246 0.74 0.39 0.05
Experiment x Trial 4 , 490 21.23 <.001* 0.42 8 , 984 36.83 <.001* 0.55 2 , 246 1.67 0.19 0.12
Drug x Experiment x Trial 4 , 490 0.36 0.84 0.05 8 , 984 1.59 0.12 0.11 2 , 246 1.96 0.14 0.13
Table S8
ANOVA FSTT
Total Sequences Correct Sequences
Effect F-value p-value cohen's d F-value p-value cohen's d
Drug (doxycycline/placebo) 1 , 245 0.37 0.54 0.04 1 , 245 0.36 0.55 0.04
Trial 20 ,4900 626.45 <.001* 1.60 20 ,4900 255.02 <.001* 1.02
Drug x Trial 20 ,4900 2.31 0.001* 0.10 20 ,4900 2.23 0.001* 0.10
Table S9
ANCOVA FSTT
Total Sequences Correct Sequences
Effect F-value p-value cohen's d F-value p-value cohen's d
Drug (doxycycline/placebo) 1 , 245 0.47 0.49 0.04 1 , 245 0.46 0.50 0.04
Trial 20 ,4900 626.45 <.001* 1.60 20 ,4900 255.02 <.001* 1.02
Drug x Trial 20 ,4900 2.31 0.001* 0.10 20 ,4900 2.23 0.001* 0.10
df
df
df
Learning
df
df
df
df

Forget me not: The effect of doxycycline on human declarative memory 6
Table S10
T-test RAVLT
df t-value p-value cohen's d hedge's g 95% CI
Supraspan (Trial A1_day0) 9.02 ± 2.22 9.47 ± 2.18 250 -1.64 0.10 -0.21 -0.21 [-1.00, 0.09]
Trial A2_day0 12.11 ± 2.00 12.36 ± 1.86 250 -1.01 0.31 -0.13 -0.13 [-0.72, 0.23]
Trial A3_day0 13.20 ± 1.73 13.44 ± 1.58 250 -1.14 0.26 -0.14 -0.14 [-0.65, 0.17]
Trial A4_day0 13.72 ± 1.47 13.80 ± 1.36 250 -0.45 0.66 -0.06 -0.06 [-0.43, 0.27]
Learning Performance (Trial A5_day0) 14.21 ± 1.22 13.99 ± 1.25 250 1.38 0.17 0.17 0.17 [-0.09, 0.52]
List B1_day0 9.06 ± 2.31 9.27 ± 2.43 250 -0.72 0.47 -0.09 -0.09 [-0.80, 0.37]
Trial A6_day0 13.67 ± 1.72 13.55 ± 1.82 250 0.53 0.59 0.07 0.07 [-0.32, 0.56]
Trial A7_day0 13.83 ± 1.72 13.93 ± 1.57 250 -0.46 0.65 -0.06 -0.06 [-0.50, 0.31]
Trial A7_day7 11.08 ± 2.87 11.35 ± 2.79 250 -0.76 0.45 -0.10 -0.10 [-0.97, 0.43]
Total Learning Performance (∑Trial A1-5_day0) 62.25 ± 7.10 63.06 ± 6.64 250 -0.93 0.36 -0.12 -0.12 [-2.51, 0.90]
Proactive Interference (Trial A1_day0 - trial B1_day0) -0.04 ± 2.14 0.20 ± 2.25 250 -0.86 0.39 -0.11 -0.11 [-0.78, 0.31]
Retroactive Interference (Trial A5_day0 - A6_day0) 0.54 ± 1.14 0.44 ± 1.27 250 0.63 0.53 0.08 0.08 [-0.20, 0.39]
Loss day0 (Trial A5_day0 -A7_day0) 0.37 ± 1.17 0.06 ± 1.20 250 2.07 0.040* 0.26 0.26 [0.01, 0.60]
Loss day7 (Trial A5_day0 -A7_day7) 3.13 ± 2.39 2.64 ± 2.44 250 1.59 0.11 0.20 0.20 [-0.12, 1.08]
Recognition day0 14.59 ± 0.93 14.67 ± 0.58 250 -0.81 0.42 -0.10 -0.10 [-0.27, 0.11]
Recognition day7 13.83 ± 1.47 13.97 ± 1.30 250 -0.77 0.44 -0.10 -0.10 [-0.48, 0.21]
Placebo
Doxycycline
Mean ± SD
Mean ± SD

Forget me not: The effect of doxycycline on human declarative memory 7
Table S11
T-test RVDLT
df t-value p-value cohen's d hedge's g 95% CI
Supraspan (Trial A1_day0) 6.04 ± 2.00 6.46 ± 1.98 250 -1.68 0.094 -0.21 -0.21 [-0.91, 0.07]
Trial A2_day0 8.85 ± 2.55 9.44 ± 2.53 250 -1.86 0.064 -0.24 -0.23 [-1.23, 0.03]
Trial A3_day0 10.42 ± 2.58 11.02 ± 2.86 250 -1.76 0.080 -0.22 -0.22 [-1.28, 0.07]
Trial A4_day0 11.08 ± 2.80 11.54 ± 2.82 250 -1.30 0.19 -0.16 -0.16 [-1.16, 0.24]
Learning Performance (Trial A5_day0) 11.56 ± 2.79 11.75 ± 2.68 250 -0.55 0.58 -0.07 -0.07 [-0.87, 0.49]
List B1_day0 7.09 ± 2.78 7.12 ± 2.67 250 -0.07 0.94 -0.01 -0.01 [-0.70, 0.65]
Trial A6_day0 10.78 ± 2.70 10.95 ± 2.83 250 -0.48 0.63 -0.06 -0.06 [-0.85, 0.52]
Trial A7_day0 11.01 ± 2.72 11.32 ± 2.94 250 -0.87 0.39 -0.11 -0.11 [-1.01, 0.39]
Trial A7_day7 8.64 ± 2.58 8.86 ± 2.89 249 -0.63 0.53 -0.08 -0.08 [-0.90, 0.46]
Total Learning Performance (∑Trial A1-5_day0) 47.93 ± 11.56 50.20 ± 11.76 250 -1.55 0.12 -0.20 -0.19 [-5.16, 0.62]
Proactive Interference (Trial A1_day0 - trial B1_day0) -1.06 ± 2.11 -0.66 ± 1.93 250 -1.56 0.12 -0.20 -0.20 [-0.90, 0.11]
Retroactive Interference (Trial A5_day0 - A6_day0) 0.78 ± 1.28 0.80 ± 1.29 250 -0.15 0.88 -0.02 -0.02 [-0.34, 0.29]
Loss day0 (Trial A5_day0 -A7_day0) 0.55 ± 1.19 0.43 ± 1.06 250 0.84 0.40 0.11 0.11 [-0.16, 0.40]
Loss day7 (Trial A5_day0 -A7_day7) 2.93 ± 2.33 2.89 ± 2.06 249 0.14 0.89 0.02 0.02 [-0.51, 0.59]
Recognition day0 14.32 ± 1.09 14.42 ± 0.90 250 -0.82 0.41 -0.10 -0.10 [-0.35, 0.14]
Recognition day7 13.80 ± 1.19 13.78 ± 1.22 250 0.16 0.88 0.02 0.02 [-0.28, 0.32]
Placebo
Doxycycline
Mean ± SD
Mean ± SD

Forget me not: The effect of doxycycline on human declarative memory 8
Table S12
T-test FSTT
Measure df t-value p-value cohen's d hedge's g 95% CI
Learning 221.16 ± 54.97 220.42 ± 45.96 245 0.12 0.91 0.01 0.01 [-11.95, 13.44]
Immediate Recall 66.36 ± 14.96 64.60 ± 13.91 245 0.95 0.34 0.12 0.12 [-1.87, 5.37]
Delayed Recall day0 76.93 ± 17.25 74.24 ± 15.66 245 1.28 0.20 0.16 0.16 [-1.44, 6.82]
Delayed Recall day7 74.35 ± 17.36 72.19 ± 15.63 245 1.03 0.31 0.13 0.13 [-1.98, 6.29]
Learning 202.62 ± 53.34 203.31 ± 42.61 245 -0.11 0.91 -0.01 -0.01 [-12.78, 11.41]
Immediate Recall 61.49 ± 14.36 59.74 ± 13.31 245 0.99 0.32 0.13 0.13 [-1.72, 5.22]
Delayed Recall day0 70.92 ± 17.85 67.65 ± 14.96 245 1.56 0.12 0.20 0.20 [-0.86, 7.39]
Delayed Recall day7 69.34 ± 18.02 66.83 ± 15.66 245 1.17 0.24 0.15 0.15 [-1.72, 6.74]
Total
Correct
Placebo
Doxycycline
Mean ± SD
Mean ± SD

Forget me not: The effect of doxycycline on human declarative memory 9
Table S13
WRS CPT
Placebo Doxycycline N z-value rank p-value
r2
Hits 0.90 0.90 172 0.33 7546.00 0.74 0.00
Misses 0.10 0.10 172 -0.38 7316.00 0.71 0.00
Table S14
WRS DSST
Score Placebo Doxycycline N z-value rank p-value
r2
Total day 0 60 63 75 -1.08 1304.00 0.28 0.02
Correct day 0 60 63 75 -0.98 1313.50 0.33 0.01
Total day 7 68 66 75 0.00 1405.50 1.00 0.00
Correct day 7 66 66 75 -0.04 1401.50 0.97 0.00
Median
Median