877954 ANP ANZJP ArticlesMcGovern et al.

Key Review

Australian & New Zealand Journal of Psychiatry

A review of the antimicrobial side 2019, Vol. 53(12) 1151­–1166

https://doi.org/10.1177/0004867419877954
DOI: 10.1177/0004867419877954

of antidepressants and its putative © The Royal Australian and

New Zealand College of Psychiatrists 2019

implications on the gut microbiome Article reuse guidelines:

sagepub.com/journals-permissions
journals.sagepub.com/home/anp

Abigail S McGovern , Adam S Hamlin and Gal Winter

Abstract
Objective: Serotonin reuptake inhibitors are the predominant treatment for major depressive disorder. In recent
years, the diversity of the gut microbiota has emerged to play a significant role in the occurrence of major depressive
disorder and other mood and anxiety disorders. Importantly, the role of the gut microbiota in the treatment of such
disorders remains to be elucidated. Here, we provide a review of the literature regarding the effects of physiologically
relevant concentrations of serotonin reuptake inhibitors on the gut microbiota and the implications this might have on
their efficacy in the treatment of mood disorders.
Methods: First, an estimation of gut serotonin reuptake inhibitor concentrations was computed based on pharmaco-
kinetic and gastrointestinal transit properties of serotonin reuptake inhibitors. Literature regarding the in vivo and in
vitro antimicrobial properties of serotonin reuptake inhibitors was gathered, and the estimated gut concentrations were
examined in the context of these data. Computer-based investigation revealed putative mechanisms for the antimicrobial
effects of serotonin reuptake inhibitors.
Results: In vivo evidence using animal models shows an antimicrobial effect of serotonin reuptake inhibitors on the gut
microbiota. Examination of the estimated physiological concentrations of serotonin reuptake inhibitors in the gastroin-
testinal tract collected from in vitro studies suggests that the microbial community of both the small intestine and the
colon are exposed to serotonin reuptake inhibitors for at least 4 hours per day at concentrations that are likely to exert
an antimicrobial effect. The potential mechanisms of the effect of serotonin reuptake inhibitors on the gut microbiota
were postulated to include inhibition of efflux pumps and/or amino acid transporters.
Conclusion: This review raises important issues regarding the role that gut microbiota play in the treatment of mood-
related behaviours, which holds substantial potential clinical outcomes for patients suffering from major depressive
disorder and other mood-related disorders.

Keywords
Depression, gut microbiota, serotonin reuptake inhibitor, antidepressants, gut microbiome, major depressive disorder,
selective serotonin reuptake inhibitors

Introduction
Antidepressants are used in the treatment of affective and
anxiety disorders, most commonly for the treatment of
major depressive disorder (MDD; Abbing-Karahagopian
et al., 2014). They are in common use worldwide, with
annual usage estimated to be ~10% in Western population
(Abbing-Karahagopian et al., 2014; Brett et al., 2017; Pratt
et al., 2017). Despite selective serotonin reuptake inhibitors
(SSRIs) being the most commonly prescribed class of anti-
depressants (Abbing-Karahagopian et al., 2014), there are
many uncertainties regarding SSRI treatment of MDD.
First, while the synaptic effect of SSRIs is instantaneous,
reductions in MDD symptomology take several weeks
School of Science and Technology, University of New England,
Armidale, NSW, Australia
Corresponding author:
Gal Winter, School of Science and Technology, University of New
England, Armidale, NSW 2351, Australia.
Email: gwinterz@une.edu.au

Australian & New Zealand Journal of Psychiatry, 53(12)

1152
(Jeon and Kim, 2016). Furthermore, only ~30–40% of the
MDD patients experience remission (Keller et al., 2007;
Rush et al., 2006) and of these patients, many (~20–40%)
relapse even with continuing treatment (Rush et al., 2006;
Sim et al., 2015). As treatment-resistant MDD presents a
substantive societal burden (Mrazek et al., 2014), a better
understanding of the factors that may interact with SSRI
mediation of antidepressant effect is imperative.
In recent years, the gut microbiota (GM) composition
has been associated with a number of illnesses, including
psychiatric disorders (Moya and Ferrer, 2016; Rojo et al.,
2017). Recent studies point towards a strong link between
the diversity of the GM and depression (Winter et al.,
2018). MDD has been associated with reductions in micro-
bial richness and diversity, as well as changes in relative
abundances of specific microbial taxa (Winter et al., 2018).
The gut microbial community is a complex ecological sys-
tem. It is composed mostly of bacteria from the Bacteroidetes
and Firmicutes phyla (90%), with the Actinobacteria,
Proteobacteria and Verrucomicrobia phyla comprising only
a small part of the GM (Eckburg et al., 2005). Other symbi-
onts of the gut include fungi (particularly yeasts) and some
viruses (Lozupone et al., 2012). Gut microbes act as a func-
tional group in a manner dependent on metabolic activity,
which is largely determined by the community composition
(Moya and Ferrer, 2016). The microbiota and associated
metabolic by-products may affect the host’s brain via a
number of pathways, including (1) vagal signalling, (2)
production of neuroactive metabolites such as catechola-
mine neurotransmitters or short chain fatty acids or (3)
indirectly via the immune system (Bonaz et al., 2018; Fung
et al., 2017; Kasubuchi et al., 2015; Rooks and Garrett,
2016; Stilling et al., 2016; Yano Jessica et al., 2015).
Gut microbiome diversity and dynamics
The dynamics of the gut microbial community are complex
due to interactions between microbial species and interac-
tions of microbes with their environment. The literature
suggests that the microbial community is drawn to a stable
equilibrium composition (Lozupone et al., 2012; Mehta
et al., 2018) whose stability is facilitated by competitive
inhibition, host regulation and adequate nutrients (Coyte
et al., 2015). However, a number of environmental and life-
style factors, including drug consumption, may lead to dys-
biosis (Cryan and Dinan, 2019; David et al., 2014; Kundu
et al., 2017; Moloney et al., 2014; Rojo et al., 2017). The
most extreme example of this is antibiotics. Even a single
course of antibiotics substantially reduce richness and
diversity of GM (Eisen et al., 2008; Isaac et al., 2017);
however, over time the GM usually progresses towards its
original equilibrium state (Rojo et al., 2017). Other phar-
maceutical agents designed to target human cells rather
than microbial cells have been shown to modulate the GM
diversity (Maier et al., 2018; Zimmermann et al., 2019).
Australian & New Zealand Journal of Psychiatry, 53(12)
ANZJP Articles
Specific examples include pharmaceuticals such as antidia-
betics (metformin; Forslund et al., 2015), proton pump
inhibitors (PPIs; Imhann et al., 2016; Jackson et al., 2016),
nonsteroidal anti-inflammatory drugs (NSAIDs; Rogers
and Aronoff, 2016) and atypical antipsychotics (AAPs;
Flowers et al., 2017). In a microbial community as diverse
as the gut ecosystem, it is logical to assume that some of the
microbes would be able to metabolise the consumed drugs
and that this will either have a growth-promoting effect or
a growth-inhibiting (antimicrobial) effect on microbes.
Indeed, microbial metabolism of over 240 pharmaceutical
drugs of all therapeutic classes has been shown in vitro
(Maier et al., 2018; Zimmermann et al., 2019). When con-
sidering the effect of drug metabolism on GM diversity, it
is important to apprehend that metabolism of the drug by
even one microbial strain will have a rolling effect on the
rest of the microbial community and will lead to changes in
the overall diversity.
As such, SSRIs have the capacity to exert a constant
antimicrobial pressure and affect community diversity,
especially considering that they are taken for long dura-
tions. For instance, in the United States, around 25% of the
individuals who used antidepressants between 2011 and
2014 had done so for a period of greater than 10 years (Pratt
et al., 2017). However, for SSRIs to affect the GM diver-
sity, they would need to be present in the gut lumen at suf-
ficient concentrations, especially in the microbial-enriched
regions such as the ascending colon (Donaldson et al.,
2015).
In this review article, we present new evidence regard-
ing in vivo antimicrobial activity of SSRIs. We then evalu-
ate the extent by which GM is exposed to SSRIs by
evaluating the gastrointestinal (GI) concentrations of
SSRIs using pharmacokinetic (PK) properties and the tem-
poral nature of their GI transit. These concentrations are
then compared with in vitro evidence assessing the effect
of SSRIs against bacteria and fungi. A computer-based
analysis is further performed to identify potential mecha-
nisms responsible for the antimicrobial effect of SSRIs.
Finally, we discuss the clinical implications of these effects
and their relevance to SSRI treatment of mood-related
disorders.
In vivo evidence for the
antimicrobial activity of SSRIs
Recent studies using animal models provide in vivo evi-
dence for the antimicrobial activity of SSRIs. Cussotto
et al. (2019) demonstrated using rats that animals treated
with fluoxetine (FLU) and escitalopram (ECIT) had an
altered microbiota composition and that FLU-treated ani-
mals showed a decrease in Deferribacteres as determined
by 16S rRNA sequencing analysis of faecal samples
(Cussotto et al., 2019). Importantly, this study was con-
ducted on healthy rat models and included administration
McGovern et al. 1153

Figure 1. Box 1 shows the daily dosage range of several common SSRIs. These include fluoxetine (FLU), sertraline (SER),
paroxetine (PRX), fluvoxamine (FVX), escitalopram (ECIT) and citalopram (CIT). Box 2 shows the percentage of the drug that
is estimated to remain, unabsorbed, in the lumen and an estimate maximal concentration as the drug passes through the ileum.
All percentages are calculated based on the literature sourced mean pharmacokinetic parameters at times that represent the
GI transit as determined experimentally using radiolabelled tablets. All concentration estimates are calculated based on taking
the maximum dose of the drug and assume complete dissolution in an experimentally determined mean fluid volume for that
region of the digestive tract. The values in brackets give some idea of the possible variation between individuals given different
gastrointestinal transit speeds; they were computed using the mean time plus or minus one standard deviation. Box 3 shows the
percentage of the drug that is unabsorbed and an estimate maximal concentration at the time that half of the drug arrives at the
colon. Concentrations are therefore calculated using half of the remaining drug. Box 4 shows the percentage of the drug that is
unabsorbed and an estimate maximal concentration 4 hours later. Concentrations assume that the entire drug has reached the
colon. Box 5 shows the percentage of a radiolabelled dose of the given SSRI excreted in the faeces in mass balance studies.

of lower doses than the recommended daily dosages used in
Figure 1. Lyte et al. (2019) administered FLU at the recom-
mended daily dosage to healthy male mice and reported a
significant, time-dependent alteration of microbial diver-
sity, based on 16S rRNA analysis of faecal samples. The
authors also report changes in body mass resulting from
FLU administration. Among the microbial genera decreased
in abundance due to FLU administration was Lactobacillus,
which has been associated with body mass regulation (Lyte
and Brown, 2018). The authors suggest that depletion of
certain microbial genera due to FLU supplementation is
partially responsible for some of the known side effects
such as weight loss (Lyte et al., 2019). An additional study
examined the effect of SSRIs on the microbiota of rats dur-
ing pregnancy and lactation (Ramsteijn et al., 2018). The
study used female rat models of maternal vulnerability
(MV), where rats are exposed to early life stress and devel-
oped a depressive-like phenotype. The authors examined
the GM diversity and metabolic capacity of MV rats treated
with FLU during pregnancy and lactation, compared to a
control group. Results from this study show that FLU
administration had a significant impact on GM diversity
and metabolic activity of MV rats, for example, lowering
the abundance of genera Prevotella, Oscillospira and
Ruminococcus. The latter genus was also identified as
decreased in abundance following administration of five
different antidepressants by Lukic et al. (2019). The study
used BALB/c mice as models as this strain is known for
having natural characteristics of exhibiting higher depres-
sive-like behaviour and anxiety compared to other strains
(Lukic et al., 2019). In follow-up experiments, BALB/c
mice were treated with the antidepressant duloxetine with

Australian & New Zealand Journal of Psychiatry, 53(12)

1154
or without a laboratory-prepared culture of Ruminococcus
flavefaciens. It was found that supplementation of R. flave-
faciens reversed the behavioural symptoms observed fol-
lowing the duloxetine administration, acting as an
anti-antidepressant. R. flavefaciens supplementation has
also been shown to induce changes in cortical gene expres-
sion leading the authors to hypothesise a mechanism for
microbial regulation of antidepressant treatment efficiency
(Lukic et al., 2019).
A limited number of observational studies in humans
have examined SSRI use as a variable when comparing
subjects’ microbiome profiles. A large observational study
(n > 2700, 113 taking SSRIs) examining the effects of 38
common diseases and 51 medications on the GM (Jackson
et al., 2018) found negative associations between a number
of taxa and SSRI use. In this study, depression did not sig-
nificantly associate with the aforementioned taxa and SSRI
use had larger correlations with a greater number of relative
abundances. In a study of 76 elderly participants on polyp-
harmacy, antidepressant use was positively correlated with
several taxa (Ticinesi et al., 2017). Investigation of the
effects of NSAIDs on the microbiome found that citalo-
pram (CIT) was significantly associated with a 32%
increase in abundance of a member of the Enterobacteriaceae
family and found that SSRI use moderated the effects of
NSAIDs (Rogers and Aronoff, 2016). Studies (Vogt et al.,
2017; Zheng et al., 2016) found no correlation between
microbiome structure and antidepressant use. However, as
these studies were not directly aimed at assessing the rela-
tionship between SSRI use and microbiome diversity, the
conclusions drawn in this regard are limited by the number
of participants and confounding factors.
Exposure of GM to SSRIs
SSRIs are lipophilic and dissolve slowly (Filipski et al.,
2013; Pawar et al., 2012; Seaman et al., 2005). In order to
assess possible concentrations of SSRIs along the gastro-
intestinal tract (GIT), evidence has been sourced from PK
and GI transit studies. We found PK research for a number
of SSRIs, including FLU, ECIT, sertraline (SER), parox-
etine (PRX), fluvoxamine (FVX) and CIT (Brown et al.,
2005; van Gorp et al., 2012; Venkatakrishnan and Obach,
2005). First-order rate constants for absorption, which
assume passive concentration-dependent absorption,
describe how rapidly a drug is absorbed as a function of
the amount of drug yet to be absorbed (Mould and Upton,
2013). This is suitable given that SSRI uptake is a largely
passive process (Pawar et al., 2012). Average absorption
rate constants for the aforementioned SSRIs were those
reported by Brown et al. (2005), van Gorp et al. (2012)
and Venkatakrishnan and Obach (2005). These were then
used to estimate the amount of drug remaining after a
given amount of time. This can be estimated assuming
that the initial amount of drug at the absorption site and
Australian & New Zealand Journal of Psychiatry, 53(12)
ANZJP Articles
the absorption fraction is known (Supplementary File 1).
Given that SSRIs are highly lipophilic and nearly com-
pletely absorbed (Hiemke and Härtter, 2000; Pawar et al.,
2012), the absorption fraction was assumed to be 1.0.
Research findings of GI drug transit times for non-disinte-
grating radiolabelled tablets (Adkin et al., 1993; Davis
et al., 1986) were used to estimate the proportion of drug
that remains at different locations in the GIT. Estimate
concentrations have been calculated and are shown in
Figure 1, and the equations used are presented in
Supplementary File 1. These concentrations are the daily
theoretical maximum ([SSRI]max) and are based on the
calculated amount remaining of the maximum daily dose
(Figure 1, box 1; Hiemke and Härtter, 2000; van Harten,
1995; Wade et al., 2011) when completely dissolved in the
estimated average fluid volumes of different sections of
the GIT. The calculations do not address potential cumu-
lative effects of long-term drug exposure.
The small intestine is home to a number of microbial
taxa (Donaldson et al., 2015). Figure 1, box 2 shows esti-
mates of remaining SSRI concentrations at a time that is
estimated as being midway through small intestinal transit,
which combined gastric emptying and half of small intesti-
nal transit (132 [±SD 60] minutes; Adkin et al., 1993;
Davis et al., 1986). Calculation of [SSRI]max used the calcu-
lated mass of remaining drug and an average ileum fluid
volume of 31 mL (Sutton, 2009). Appreciable amounts of
SSRIs are predicted as being present. However, greater pro-
portions of FLU, SER, PRX and FVX than ECIT and CIT
are predicted as remaining (undissolved), which is consist-
ent with these drugs taking less time to reach peak plasma
concentration (De Vane et al., 2002; Hiemke and Härtter,
2000; Jovanović et al., 2006; Qamar and Safila, 2015).
As the colon is home to the most abundant microbial pop-
ulation, likely SSRI exposure was estimated for two points:
arrival time and after 4 hours (Figure 1, boxes 3 and 4).
Concentrations were calculated based on a mean colonic
fluid volume of 11 mL (Schiller et al., 2005). Box 3 esti-
mates used the mean (±SD) time taken for 50% of a drug
to arrive at the colon of 210 (±111) minutes (Adkin et al.,
1993) and 50% of the mass of the remaining drug. Box 4
shows estimates for 4 hours later and assumes that the
entirety of the remaining drug is present in the colon.
Estimates indicate that substantial amounts of FLU, SER,
PRX and FVX likely remain in the colon for at least 4 hours,
ample time for these SSRIs to exert an antimicrobial effect.
Finally, Figure 1, box 5 shows the proportion of excretion
of SSRIs and their metabolites in the faeces, as identified
by mass balance studies (Hiemke and Härtter, 2000;
Lemberger et al., 1985; Qamar and Safila, 2015).
The calculated [SSRI]max should be interpreted with cau-
tion under the following limitations. (1) The calculations
assume total dissolution. However, the solubilities of
hydrochloride salts of even the most lipophilic SSRIs SER
and FLU (4240 and 1100 mg/L, respectively; Childs et al.,
McGovern et al.
2004; Deak et al., 2006) would permit concentrations close
to estimates in Figure 1. In the small intestine, the dissolu-
tion of SSRIs would be facilitated by bile salts (Mithani
et al., 1996) and colonic dissolution of SSRIs would likely
be facilitated by colonic mechanical forces, phospholipids,
other fats and lectins (Shameem et al., 1995; Sutton, 2009).
(2) It should be noted that the PKs of SSRIs (De Vane et al.,
2002; Hiemke and Härtter, 2000; Jovanović et al., 2006)
and GI transit of medications (Adkin et al., 1993; Davis
et al., 1986; Schiller et al., 2005) are highly variable. The
actual variability of gut concentrations of SSRIs in the pop-
ulation will far exceed the ranges reported in Figure 1.
Overall, the main finding of this analysis is that the
microbial community present in both the small intestine
and the colon is exposed to substantial concentrations of
SSRIs for a period of a minimum 4 hours on a daily basis.
Given that the estimated average doubling time for microbes
in the mammalian gut is ~3 hours (Myhrvold et al., 2015),
it is likely that this exposure will have a long-term effect on
the microbial population, in particular when considering
that SSRIs are taken daily and for a long period of time.
Furthermore, as demonstrated in Figure 1, box 5, metabo-
lites of SSRIs are also passing through the GIT and may
exert additional influence on the microbial population.
Taken together, the information provided here offers cru-
cial context for the interpretation of in vitro assessments of
the antimicrobial activity of SSRIs and their potential mod-
ulation of the GM and implications for treatment efficacy.
In vitro evidence for antimicrobial
activity of SSRIs
Research analysed and methods
The potential for the estimated concentrations of SSRIs in
the gut to confer antimicrobial effect was estimated by
reviewing the literature regarding antimicrobial effect of
SSRIs, as estimated by in vitro experiments. A number of in
vitro assays of the antimicrobial activity of SSRIs have
been conducted over the last two decades. Much of this
research involved determination of minimum inhibitory
concentrations (MICs), which are considered to be the most
appropriate and comparable means of determining antimi-
crobial activity of a compound against a given strain, as
determination involves adherence to a standardised proto-
col (Andrews, 2001). As such, these were targeted in litera-
ture searches. Searches were conducted on Google Scholar
and search terms included ‘Antimicrobial’, and the names
of the SSRIs ‘citalopram’, ‘escitalopram’, ‘fluoxetine’,
‘sertraline’, ‘paroxetine’ and ‘fluvoxamine’. The top 100
results for each search were examined for MICs of SSRIs.
The literature was also searched for research regarding
antifungal activity – particularly the determination of mini-
mum fungicidal concentrations (MFCs), the fungal equiva-
lent to MICs (Lass-Flörl et al., 2001).
1155
The antibacterial activity of a drug can be described in
terms of its MIC against a given species and the concentra-
tions to which the species is likely exposed. According to the
Clinical Lab Standards Institute (CLSI), bacteria are classi-
fied as (1) susceptible to an agent if achievable concentra-
tions are above the MIC or (2) intermediate in susceptibility
when only some isolates or strains are susceptible. When
attainable concentrations fall below the MIC, microbes are
said to be (3) resistant (Patel, 2017). This terminology will
be used when describing the efficacy of SSRIs as antimicro-
bials. The terms resistant, intermediate and susceptible are
used in accordance with concentrations reported in Figure 1.
Many studies that determined MICs reported that standard
protocols were enforced (Ayaz et al., 2015b; Bohnert et al.,
2011) although some did not provide this information
(Hendricks, 2007; Munoz-Bellido et al., 2000).
Antibacterial activity
The MICs for SSRIs sourced from the literature are shown
in Table 1. For each SSRI, the table shows the range of
MICs for tested species of Gram-positive, Gram-negative
and acid-fast bacteria. The table indicates the number of
strains upon which this was determined. The tested strains
were either reference strains or clinical isolates.
Although these strains do not necessarily represent gut
microbes, a few general comments can be made with
regards to the antimicrobial activity of SSRIs and the theo-
retical concentrations calculated in Figure 1. First and most
important, in the gut environment, many of the tested spe-
cies would be classified as susceptible or intermediate.
Importantly, this could remain the case for several hours.
The research therefore indicates that some SSRIs may exert
antimicrobial activity against gut microbes for several
hours each day. The concentrations estimated in Figure 1
indicate that the MICs of SER, FLU, PRX and FVX for a
number of bacterial species would be possible in the ileum
and ascending colon. The MICs listed in Table 1 indicate
that there is a substantial difference between different
strains and species in susceptibility to SSRIs and show that
SER and FLU may have stronger inhibitory effects on
Gram-positive bacteria than Gram-negative bacteria.
SSRIs differ in the degree to which they inhibit bacterial
growth. Examination of MICs for a variety of strains
between different SSRIs indicates that SER and FLU have
the strongest antimicrobial activity, followed by PRX and
FVX, then ECIT and CIT (Bohnert et al., 2011; Hendricks,
2007; Kalaycı et al., 2014; Munoz-Bellido et al., 2000). For
SER, MIC ranges for all species fall below the mean
[SSRI]max estimate 4 hours after colonic arrival, indicating
that many microbes would be susceptible. This was further
supported by research involving disc diffusion assays. Disc
diffusion assays involve placement of a drug-coated disc
onto inoculated agar and incubation, thus allowing the size
of zone of microbial growth inhibition to be measured. The
Australian & New Zealand Journal of Psychiatry, 53(12)
1156 ANZJP Articles

Table 1. Minimum inhibitory concentrations (MICs) for different SSRIs against various bacteria.

SSRI Microbe MIC (mg/L) Reference

(no. strains or isolates tested) MIC

Sertraline
Gram-positive
Staphylococcus aureus (9) 10–60 Ayaz et al. (2015b)
(1) 4 Kalaycı et al. (2014)
(32) 4–32 Munoz-Bellido et al. (2000)
(1) 6.25 Kumar et al. (2015)
(10) 8–16 Hendricks (2007)
(1) 160 Kruszewska et al. (2004)
MRSA (1) 32 Kalaycı et al. (2014)
Staphylococcus epidermidis (4)16 Hendricks (2007)
(1) 4 Kalaycı et al. (2014)
(32) 4–16 Munoz-Bellido et al. (2000)
VRE (1) 8 Kalaycı et al. (2014)
Enterococcus faecalis (5) 16 Hendricks (2007)
(32) 2–16 Munoz-Bellido et al. (2000)
(1) 16 Kalaycı et al. (2014)
Streptococcus pyogenes (8)16 Hendricks (2007)
(32) 16–64 Munoz-Bellido et al. (2000)
Streptococcus pneumoniae (7) 16 Hendricks (2007)
(32) 16–64 Munoz-Bellido et al. (2000)
Bacillus subtilis (1) 8 Kalaycı et al. (2014)
(1) 6.25 Kumar et al. (2015)
Streptococcus agalactiae (32) 16–64 Munoz-Bellido et al. (2000)
Clostridium perfringens (32) 32–128 Munoz-Bellido et al. (2000)
Clostridium difficile (32) 32–128 Munoz-Bellido et al. (2000)
Corynebacterium group D2 2–8 Munoz-Bellido et al. (2000)
Gram-negative
Escherichia coli (6) 8.6–68.5 69
(16) 10–100 Ayaz et al. (2015b)
(1) 4 Kalaycı et al. (2014)
(32) 64–128 Munoz-Bellido et al. (2000)
(1) 50 Kumar et al. (2015)
Pseudomonas aeruginosa (1) 60 Ayaz et al. (2015b)
Proteus vulgaris (1) 16 Kalaycı et al. (2014)
(32) 128–512 Munoz-Bellido et al. (2000)
Klebsiella pneumoniae (1) 16 Kalaycı et al. (2014)
(32) 64–256 Munoz-Bellido et al. (2000)
Acinetobacter baumannii (1) 4 Kalaycı et al. (2014)
16
Yersinia enterocolitica (1) 16 Kalaycı et al. (2014)
Proteus mirabilis (32) 32–512 Munoz-Bellido et al. (2000)
Enterobacter cloacae (32) 64–128 Munoz-Bellido et al. (2000)

(continued)

Australian & New Zealand Journal of Psychiatry, 53(12)

McGovern et al. 1157

Table 1. (Continued)

SSRI Microbe MIC (mg/L) Reference

(no. strains or isolates tested) MIC

Citrobacter freundii (32) 8–64 Munoz-Bellido et al. (2000)

Serratia marcescens
Morganella morganii
  Salmonella sp.
Haemophilus influenzae
  Moraxella catarrhalis
  Campylobacter jejuni
  Bacteroides fragilis
  Prevotella spp.
  Brucella spp.
Acid-fast
Mycobacterium tuberculosis
(32) 256–(>512) Munoz-Bellido et al. (2000)
(32) 256–512 Munoz-Bellido et al. (2000)
(32) >512 Munoz-Bellido et al. (2000)
(32) 8–64 Munoz-Bellido et al. (2000)
(32) 4–16 Munoz-Bellido et al. (2000)
(32) 2–32 Munoz-Bellido et al. (2000)
(32) 16–32 Munoz-Bellido et al. (2000)
(32) 16–64 Munoz-Bellido et al. (2000)
(32) 8–16 Munoz-Bellido et al. (2000)
(1) 2 Kalaycı et al. (2014)
(1) 1.6 Kumar et al. (2015)

Fluoxetine
Gram-positive
S. aureus (1) 4 Kalaycı et al. (2014)
(1) 12.5 Kumar et al. (2015)
MRSA (1) 128 Kalaycı et al. (2014)
S. epidermidis (1) 32 Kalaycı et al. (2014)
VRE (1) 16 Kalaycı et al. (2014)
E. faecalis (1) 32 Kalaycı et al. (2014)
B. subtilis (1) 4 Kalaycı et al. (2014)
(1) 12.5 Kumar et al. (2015)
Corynebacterium group D2 8–32 Munoz-Bellido et al. (2000)
Gram-negative
E. coli (6) 17.3–277 Bohnert et al. (2011)
(1) 16 Kalaycı et al. (2014)
(1) 100
P. vulgaris (1) 32 Kalaycı et al. (2014)
K. pneumonia (1) 32 Kalaycı et al. (2014)
A. baumannii (1) 32 Kalaycı et al. (2014)
Y. enterocolitica (1) 64 Kalaycı et al. (2014)
Acid-fast
M. tuberculosis (1) 8 Kalaycı et al. (2014)

Citalopram
Gram-positive
S. aureus (2) 4000–5000 Ayaz et al. (2015a)
(1) 16,000 Kruszewska et al. (2006)
(1) 13,000 Kruszewska et al. (2008)
(10) >256 Hendricks (2007)
(1) 256 Kalaycı et al. (2014)

(continued)

Australian & New Zealand Journal of Psychiatry, 53(12)

1158 ANZJP Articles

Table 1. (Continued)

SSRI Microbe MIC (mg/L) Reference

(no. strains or isolates tested) MIC

S. epidermidis (4) >256 Hendricks (2007)

E. faecalis (1) 7000 Ayaz et al. (2015a)
(5) >256
S. pyogenes (8) >256 Hendricks (2007)
S. pneumonia (7) >256 Hendricks (2007)
B. subtilis (1) 256 Kalaycı et al. (2014)
Gram-negative
E. coli (2) >800 µM Bohnert et al. (2011)
(2) 5000–7000 Ayaz et al. (2015a)
(1) 16,000 Kruszewska et al. (2006)
(1) 7000 Kruszewska et al. (2008)
(1) 256 Kalaycı et al. (2014)
P. aeruginosa (2) 4000–6000 Ayaz et al. (2015a)
Salmonella typhi (1) 6000 Ayaz et al. (2015a)
K. pneumonia (1) 6000 Ayaz et al. (2015a)
(1) 512 Kalaycı et al. (2014)
P. mirabilis (1) 6000 Ayaz et al. (2015a)
A. baumannii (1) 256 Kalaycı et al. (2014)

Fluvoxamine
Gram-positive
S. aureus (1) 4000 Kruszewska et al. (2010)
S. epidermidis (1) 64 Kalaycı et al. (2014)
B. subtilis (1) 64 Kalaycı et al. (2014)
Gram-negative
E. coli (1) 32 Kalaycı et al. (2014)
(1) 2000 Kruszewska et al. (2010)
P. aeruginosa (1) 10,000 Kruszewska et al. (2010)
P. vulgaris (1) 64 Kalaycı et al. (2014)
K. pneumoniae (1) 32 Kalaycı et al. (2014)
A. baumannii (1) 128 Kalaycı et al. (2014)
Y. enterocolitica (1) 256 Kalaycı et al. (2014)

Femoxetine
Gram-positive
S. aureus (1) 1000 Kumar et al. (2015)

Paroxetine
Gram-positive
S. aureus (1) 128 Kalaycı et al. (2014)
(1) 25 Kumar et al. (2015)
MRSA (1) 128 Kalaycı et al. (2014)
S. epidermidis (1) 32 Kalaycı et al. (2014)
VRE (1) 32 Kalaycı et al. (2014)

(continued)

Australian & New Zealand Journal of Psychiatry, 53(12)

McGovern et al. 1159

Table 1. (Continued)

SSRI Microbe MIC (mg/L) Reference

(no. strains or isolates tested) MIC
E. faecalis (1) 64 Kalaycı et al. (2014)
B. subtilis (1) 256 Kalaycı et al. (2014)
(1) 25 Kumar et al. (2015)
Corynebacterium group D2 4–32 Munoz-Bellido et al. (2000)
Gram-negative
E. coli (6) 18.3–293 Bohnert et al. (2011)
(1) 128 Kalaycı et al. (2014)
(1) 100
P. vulgaris (1) 32 Kalaycı et al. (2014)
K. pneumoniae (1) 64 Kalaycı et al. (2014)
A. baumannii (1) 64 Kalaycı et al. (2014)
Y. enterocolitica (1) 64 Kalaycı et al. (2014)
Acid-fast
M. tuberculosis (1) 8 Kalaycı et al. (2014)

Escitalopram
Gram-positive
B. subtilis (1) 256 Kalaycı et al. (2014)
Gram-negative
P. vulgaris (1) 128 Kalaycı et al. (2014)
K. pneumoniae (1) 256 Kalaycı et al. (2014)
A. baumannii (1) 128 Kalaycı et al. (2014)
Y. enterocolitica (1) 512 Kalaycı et al. (2014)

SSRI: serotonin reuptake inhibitor; MIC: minimal inhibitory concentration; MRSA: methicillin-resistant Staphylococcus aureus; TCA: tricyclic
antidepressant; VRE: vancomycin-resistant Enterococcus.

authors demonstrated differences in the number of organ-
isms inhibited by different SSRIs. They found that SER,
FLU and PRX affected the most species (12/13), followed
by FVX (7/13), and CIT and ECIT (5/13; Kalaycı et al.,
2014).
The variability in MICs reported for different strains and
species may be an important factor in the gut, as intermicrobe
differences in inhibition would facilitate a change in commu-
nity structure. Taxonomically, some assayed bacteria
belonged to genera that are commonly associated with dys-
biosis, including Enterococcus, Streptococcus, Lacto­bacillus,
Clostridium, Bifidobacterium, Prevotella, Bac­teroides and
Escherichia (Rojo et al., 2017). In regards to specific spe-
cies, MICs indicate that Bacillus subtilis and Bacteroides
fragilis, which are commensals implicated in development of
gut-associated lymphoid tissue (GALT; Rhee et al., 2004),
could be inhibited by SER or FLU. Gut concentrations may
also inhibit Escherichia coli but be less inhibitory to
Clostridium species, both of which may be found in the gut
(Atarashi et al., 2011; Eckburg et al., 2005). As Candida spp.
have important roles in gut microbial ecology, antifungal
activity of SSRIs may also affect gut ecology. Particularly,
Candida albicans interacts both with the host – through pro-
duction of immunomodulatory compounds – and bacterial
symbionts (Huffnagle and Noverr, 2013).
Antifungal activity
SSRIs exhibit some antifungal activity, which is generally
weaker than their antibacterial activity. In assays of antifun-
gal activity, fewer species and strains have been tested,
from fewer genera. The MFCs found were tabulated and
can be found in Table 2. The inhibitory strength of different
SSRIs follows a similar trend in fungi as in bacteria. Studies
by Lass-Flörl et al. (2001) and Kalaycı et al. (2014) indicate
that SER, FLU and PRX may have antifungal activity
against some strains at the estimated concentrations. The
overall effect of SSRIs on fungal physiology is not fully
understood; however, this could be an important direction
for future studies.

Australian & New Zealand Journal of Psychiatry, 53(12)

1160 ANZJP Articles

Table 2. Minimum fungicidal concentrations for different SSRIs against various species of fungi.

SSRI Microbe MFC(mg/L) Reference

(no. strains or isolates tested) MFC

Sertraline
Aspergillus fumigatus (10) 7–327 Lass-Flörl et al. (2001)a
(5) 80–100 Ayaz et al. (2015b)
Aspergillus niger (5) 20–80 Ayaz et al. (2015b)
(1) 64 Kalaycı et al. (2014)
Aspergillus flavus (5) 60–80 Ayaz et al. (2015b)
(9)14–475 Lass-Flörl et al. (2001)a
Aspergillus terreus (10) 29–237 Lass-Flörl et al. (2001)a
Candida parapsilosis (1) 59 Lass-Flörl et al. (2001)a
Candida albicans (1) 128 Kalaycı et al. (2014)
(1) 25 Kumar et al. (2015)
Fusarium solani (1) 80 Ayaz et al. (2015b)

Fluoxetine
A. fumigatus (10) 78–1250 Lass-Flörl et al. (2001)a
A. flavus (9) 156–2500 Lass-Flörl et al. (2001)a
A. terreus (10) 59–2500 Lass-Flörl et al. (2001)a
C. parapsilosis (1) 78–156 Lass-Flörl et al. (2001)a
C. albicans

Paroxetine
A. fumigatus (10) 62–2500 Lass-Flörl et al. (2001)a
A. flavus (9) 62–1000 Lass-Flörl et al. (2001)a
A. terreus (10) 500–1000 Lass-Flörl et al. (2001)a
C. parapsilosis (1) 500 Lass-Flörl et al. (2001)a
C. albicans (1) 25 Kumar et al. (2015)

Citalopram
A. fumigatus (10) 312–1250 Lass-Flörl et al. (2001)a
A. flavus (9) 312–2500 Lass-Flörl et al. (2001)a
A. terreus (10) 312–1250 Lass-Flörl et al. (2001)a
C. parapsilosis (1) 1250 Lass-Flörl et al. (2001)a
C. albicans

SSRI: serotonin reuptake inhibitor; MIC: minimal inhibitory concentration; MFC: minimum fungicidal concentrations.
a8-hour inhibition.

Antimicrobial mode of action
Intracellular effects
In humans, SSRIs exert their effects on extracellular seroto-
nin levels by competitively inhibiting the serotonin trans-
porter (SERT or solute carrier family 6 member 4
[SCL6A4]). However, the mechanism of action on microbes
is less understood. As discussed, FLU and SER may have
stronger antimicrobial action than other SSRIs. This could
be informative with respect to mechanism. As determined
by octanol–water partition coefficients (P), SER and FLU
are more hydrophobic (logP, 5.1, 4.1) than ECIT, CIT, PRX
and FVX (logP, 3.5, 3.5, 3.3, 3.6; Kim et al., 2016). SER
and FLU may undergo passive diffusion across the phos-
pholipid membrane with more ease, allowing interaction
with cellular machinery (Hancock and Bell, 1989). This
hypothesis is consistent with research indicating that the
optimum lipophilicity for an antimicrobial is logP 5.5
(Wernert et al., 2004). One possible intracellular effect is
increased production of reactive oxygen species (ROS) and
mutagenesis, as has been demonstrated with FLU and
E. coli than FLU (Jin et al., 2018; Figure 2). While ROS

Australian & New Zealand Journal of Psychiatry, 53(12)

McGovern et al. 1161

Figure 2. Potential antimicrobial mechanisms of SSRIs. SSRIs may gain access to the cytoplasm via passive diffusion. They may
inhibit microbial efflux systems and reduce excretion of antimicrobials, thereby reducing antimicrobial resistance. They may
also inhibit sodium-dependent symporters, including amino acid transporters and transporters for which substrates are not
yet known. This may affect nutrient intake and excretion of waste products and metabolites. SSRI, selective serotonin reuptake
inhibitors; NSS, neurotransmitter sodium symporter family; ROS, reactive oxygen species.

may mediate the effects of SSRIs on growth, the exact
mechanisms remain unclear.
Inhibition of efflux pumps
SSRIs have been shown to inhibit microbial efflux pumps,
which contribute to antimicrobial resistance (Munoz-
Bellido et al., 2000). Efflux systems extrude antimicrobials
and other drugs from the cytoplasm. While the efflux sys-
tems are mainly associated with the expulsion of xenobiot-
ics, other physiological roles, such as in biofilm formation
and virulence of some enteric pathogens, have been pro-
posed (Opperman and Nguyen, 2015). Mechanisms of
efflux inhibition include competitive and non-competitive
inhibition and clogging of the pump’s external pore (Askoura
et al., 2011). Both SER and PRX were found to inhibit efflux
pump activity in E. coli, Pseudomonas aeruginosa and
Staphylococcus aureus at concentrations below MIC
(Bohnert et al., 2011; Kaatz et al., 2003). Interestingly, the
observed SSRI efflux inhibition was not associated with
direct growth or survival inhibition (Bohnert et al., 2011).
Nonetheless, efflux inhibition may act in concert to promote
the activity of other antimicrobials (Bohnert et al., 2011). A
number of studies have assessed the synergism between
SSRIs and various classes of antibiotics and have found
that, in some strains, the addition of SSRIs substantially
reduced the antibiotic MICs by 2–8-fold (Ayaz et al., 2015b;
Bohnert et al., 2011). This is suggestive of an indirect anti-
microbial mechanism of SSRIs (Figure 2).
Sodium-dependent transport
The action of antidepressants on SERT may offer insights into
the potential mechanism of antimicrobial function. SSRIs
competitively inhibit SERT by binding at the serotonin bind-
ing site and preventing conformational change (Grouleff
et al., 2016; Singh et al., 2007). Considering the large homol-
ogy between SERT and bacterial amino acid transporters
(Grouleff et al., 2016; Singh, 2014; Yamashita et al., 2005)
and the role these play in metabolism (Burkovski and Krämer,
2002), it is plausible that SSRIs exert similar effects in
microbes and prevent the activity of amino acid transporters,
leading to a deleterious effect on growth and survival.
In order to ascertain the scope of potential targets for
SSRIs in bacteria, a BLAST search (Altschul et al., 1990)
for the 100 most similar bacterial proteins in National
Center for Biotechnology Information (NCBI) protein
database (Coordinators, 2018) was conducted for homo-
logues of SERT. A total of 100 orthologues (E < 0.05) were
identified across a broad variety of bacterial species. All of
which were uncharacterized putative sodium-dependent
transporters. A multiple sequence alignment (Edgar, 2004;
Kumar et al., 2016) indicated that several important resi-
dues in LeuT/SERT were conserved in all of these proteins.
These included R30/104 and D404/E493, which are gating
residues, critical in the transport cycle (Grouleff et al.,
2016) and S355/438, a determinant of antidepressant bind-
ing affinity (Andersen et al., 2009). Therefore, there may be
a large number of functionally similar transporters to which
SSRIs could bind to and inhibit. As the substrates of these

Australian & New Zealand Journal of Psychiatry, 53(12)

1162
transporters have not yet been determined, inhibition may
have any number of metabolic consequences. Sodium-
dependent transport needs to be investigated with respect to
SSRI antimicrobial properties (Figure 2).
Sub-inhibitory concentrations
Given the wide variation in the efficacy of SSRIs between
species and strains, and that concentrations of SSRIs in the
gut will fluctuate, the effect of sub-inhibitory concentra-
tions (<MIC) should be considered. Research has indicated
that most antimicrobials have a biphasic effect across a sub-
inhibitory concentration range (Davies et al., 2006). To this
effect, low concentrations are often stimulatory to growth
and high concentrations become progressively more inhibi-
tory (Davies et al., 2006). The stimulatory, or hormetic,
effect on microbes is thought to be due to pro-survival
changes that at low concentrations outweigh negative
effects. This involves transcriptional alteration of around
5–10% of transcripts including upregulation of pro-survival
and pro-growth factors (e.g. heat-shock proteins and pro-
teins involved in replication; Davies et al., 2006; Yim et al.,
2007). Transcriptional changes may occur at as little as 1%
of the MIC (Yim et al., 2007). It is possible that at gut con-
centrations of SSRIs some species would undergo hormetic
changes while others would be inhibited.
Potential frameworks for SSRIs
modulation of the gut microbial
community
In addition to their direct inhibitory activity, SSRIs may
affect gut ecology indirectly by enhancing the activity of
other antimicrobial agents, for example, by inhibition of
Efflux pump. Antibiotics are in common use, and evidence
indicates that a baseline exposure to antimicrobials in house-
hold products is sufficient to affect the GM (Yee and Gilbert,
2016). Therefore, SSRIs may synergize with these antimi-
crobials to have a greater effect on microbial diversity. In
addition, SSRIs may also interact with gut microbial ecology
through their metabolic by-products, which are excreted into
the faeces (Hiemke and Härtter, 2000). Furthermore, gut
microbes may themselves metabolise SSRIs. For instance,
FLU has been found to be metabolised by a number of gut
microbe strains in vitro (Zimmermann et al., 2019). While
the effect of SSRI metabolites on gut microbes have not been
characterised, they could impact the dynamic metabolic net-
work in the gut and alter GM composition and function.
Another avenue to consider as a potential effect of SSRIs
on the GM may involve a host-mediated component. SSRIs
may directly interact with gut cells and/or immune cells to
modulate the gut environment (Mawe and Hoffman, 2013).
One application might be altered availability of gut seroto-
nin which could affect a large proportion of enteric neurons
and alter peristalsis and reflex electrolyte secretion
Australian & New Zealand Journal of Psychiatry, 53(12)
ANZJP Articles
(Schemann and Neunlist, 2004). Changes to motility and
the mucosal environment can subsequently affect the GM
(Carabotti et al., 2015). SSRIs also affect the immune sys-
tem. SERT expression is increased in activation of both
dendritic and B cells, and SSRIs have been shown to stimu-
late natural killer (NK) cells (Ahern, 2011), inhibit B cells
apoptosis (Arreola et al., 2015) and promote the Th1, over
the Th2, response (Martino et al., 2012). This may have
implications for gut microbes. Interaction with the immune
system and gut modulation may therefore be another means
by which SSRIs may manipulate the microbiome.
Finally, SSRI treatment of anxiety and mood disorders
may indirectly impact the GM through brain to gut com-
munication. As research has indicated that stress and
depression have implications for microbial diversity
(Winter et al., 2018), it is probable that alleviation of symp-
toms would also impact the microbiota. Reduced stress
may result in decreased sympathetic tone, which can affect
the GM, as GALT is directly sympathetically innervated
(Powell et al., 2017). Furthermore, some studies have found
that SSRIs modulate hypothalamic pituitary adrenal (HPA)
axis activity. Research has identified both positive and neg-
ative changes in plasma and salivary adrenocorticotropic
hormone (ACTH) and cortisol following treatment with
different SSRIs (Glintborg et al., 2018; Nikisch et al., 2005)
– although this is not always observed (Frase et al., 2018).
Cortisol modulation of the immune system may also facili-
tate a host-mediated effect of SSRIs on the GM (Powell
et al., 2017).
Implications for treatment efficacy
The effect of SSRIs on GM community composition may
have implications for their treatment efficacy. While there
are no experimental data yet available to support this, there is
evidence that antimicrobial-induced dysbiosis can have con-
sequences for the brain and behaviour, in rodents. Most sig-
nificantly, several studies identified increased depression-like
behaviour in adult-treated rats and mice (Guida et al., 2018;
Hoban et al., 2016) and decreased anxiety-like behaviour in
prenatally- and adolescent-treated mice (Desbonnet et al.,
2015; Leclercq et al., 2017). It is important to note that
some antimicrobials do have direct action on the brain
(Macedo et al., 2017). However, Guida et al. (2018) did
control for these direct effects by using an intraperitoneal
injection control group. These studies show that inducing
changes in GM composition can influence anxiety and
depression-like behaviour and thus provide an empirical
basis for investigating the effect of SSRIs on the GM and
how this interacts with treatment efficacy.
There are a number of ways in which SSRI-mediated
dysbiosis could theoretically interfere with treatment effi-
cacy. For instance, in some rodent studies antibiotic-
induced dysbiosis has been shown to reduce hippocampal
levels of brain-derived neurotrophic factor (BDNF;
McGovern et al.
Desbonnet et al., 2015; Frohlich et al., 2016; Guida et al.,
2018), which is known to promote adult hippocampal neu-
rogenesis (Ortega-Martinez, 2015). Given the evidence that
adult hippocampal neurogenesis may be necessary for the
action of a number of antidepressants (Sahay and Hen,
2007), GM-mediated changes in BDNF may be one means
by which the antimicrobial action of SSRIs could affect
treatment. Gut-mediated changes in BDNF may be brought
about by vagal nerve signalling (O’Leary et al., 2018) or
through changes in the production of metabolites such as
butyrate, which is known to increase BDNF and exert anti-
depressant-like activity (Stilling et al., 2016). In addition,
there could be gut-mediated effects on the serotonin sys-
tem. Antibiotic-induced dysbiosis has been shown to reduce
hippocampal serotonin (Hoban et al., 2016), and GM com-
position has implications for serotonin metabolism
(O’Mahony et al., 2015). Given that SSRIs are milder anti-
microbials than antibiotics, SSRI-mediated GM changes
could, theoretically, interact positively or negatively with
drug efficacy, depending on which microbes were advan-
taged or disadvantaged by the selection pressure.
Conclusion and future directions
In conclusion, there is substantial evidence that SSRIs may
affect the community structure of the gut microbiome.
While there are no direct investigations of SSRI-induced gut
dysbiosis in humans, there is a considerable amount of
research investigating the antimicrobial properties of SSRIs.
When this information is framed within the context of the
amount of SSRIs likely remaining in the gut, assessed based
on findings of PK and GI transit studies, it is probable that
many microbes may be susceptible to SSRIs. Specifically,
SER, FLU and PRX may act as antimicrobials in the gut;
however, differences in absorption and GI transit could lead
to interpersonal variability in degree of susceptibility to dif-
ferent drugs. Given the complex interactions that determine
GM community structure, it is likely that these SSRIs would
act as a persistent pressure, altering the stable state of the gut
microbiome. They may cause both increases and decreases
in the relative abundances of different gut taxa; given inter-
individual variability in gut microbiome structure, this effect
would likely differ between individuals.
If SSRIs impact the GM, research suggests that they
have the potential to interfere with treatment efficacy.
Studies of the behavioural and neurobiological effects of
antibiotic-induced dysbiosis give insight into potential
means by which this may occur. Induced dysbiosis may
affect the serotoninergic system (Desbonnet et al., 2015),
which is the initial target of SSRIs, as well as adult hip-
pocampal neurogenesis, which is thought to be an indi-
rect target of SSRIs (Möhle et al., 2016). Given the
complexity of the gut microbiome, there could be poten-
tial for both positive and negative effects on treatment
efficacy. Due to differences in the microbiomes of
1163
individuals, and in the magnitude and nature of the effects
of antimicrobial effects of SSRIs, this may account for
some of the interpersonal variabilities in treatment effi-
cacy with SSRIs.
Here, we have highlighted the need for research inves-
tigating the role of the gut–brain axis in the sporadic and
unpredictable treatment efficacy of SSRIs. Future research
should attempt to establish the antimicrobial activity of
SSRIs against gut microbes both in vitro and in vivo and
determine the effect this activity has on treatment efficacy.
Considering the significant role that gut microorganisms
are discovered to play in human physiology, it is predicted
by the authors that this type of investigation will form a
routine assessment of drug efficacy in the future.
Acknowledgements
The authors thank Professor Natkunam Ketheesan for critically
reviewing the manuscript.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with
respect to the research, authorship and/or publication of this
article.
Funding
The author(s) received no financial support for the research,
authorship and/or publication of this article.
ORCID iDs
Abigail S McGovern https://orcid.org/0000-0002-9641-2768
Gal Winter https://orcid.org/0000-0003-3789-395X
Supplemental material
Supplemental material for this article is available online.
References
Abbing-Karahagopian V, Huerta C, Souverein PC, et al. (2014)
Antidepressant prescribing in five European countries: Application
of common definitions to assess the prevalence, clinical observa-
tions, and methodological implications. European Journal of Clinical
Pharmacology 70: 849–857.
Adkin D, Davis S, Sparrow R, et al. (1993) Colonic transit of different
sized tablets in healthy subjects. Journal of Controlled Release 23:
147–156.
Ahern GP (2011) 5-HT and the immune system. Current Opinion in
Pharmacology 11: 29–33.
Altschul SF, Gish W, Miller W, et al. (1990) Basic local alignment search
tool. Journal of Molecular Biology 215: 403–410.
Andersen J, Taboureau O, Hansen KB, et al. (2009) Location of the anti-
depressant binding site in the serotonin transporter: Importance of
Ser-438 in recognition of citalopram and tricyclic antidepressants.
Journal of Biological Chemistry 284: 10276–10284.
Andrews JM (2001) Determination of minimum inhibitory concentra-
tions. Journal of Antimicrobial Chemotherapy 48: 5–16.
Arreola R, Becerril-Villanueva E, Cruz-Fuentes C, et al. (2015)
Immunomodulatory effects mediated by serotonin. Journal of
Immunology Research 2015: 354957.
Australian & New Zealand Journal of Psychiatry, 53(12)
1164
Askoura M, Mattawa W, Abujamel T, et al. (2011) Efflux pump inhibitors
(EPIs) as new antimicrobial agents against Pseudomonas aeruginosa.
Libyan Journal of Medicine 6: 5870.
Atarashi K, Tanoue T, Shima T, et al. (2011) Induction of colonic regu-
latory T cells by indigenous Clostridium species. Science 331:
337–341.
Ayaz M, Subhan F, Ahmed J, et al. (2015a) Citalopram and venlafaxine
differentially augments antimicrobial properties of antibiotics. Acta
Poloniae Pharmaceutica: Drug Research 72: 1269–1278.
Ayaz M, Subhan F, Ahmed J, et al. (2015b) Sertraline enhances the activ-
ity of antimicrobial agents against pathogens of clinical relevance.
Journal of Biological Research 22: 4.
Bohnert JA, Szymaniak-Vits M, Schuster S, et al. (2011) Efflux inhibition
by selective serotonin reuptake inhibitors in Escherichia coli. Journal
of Antimicrobial Chemotherapy 66: 2057–2060.
Bonaz B, Bazin T and Pellissier S (2018) The vagus nerve at the interface
of the microbiota-gut-brain axis. Frontiers in Neuroscience 12: 49.
Brett J, Karanges EA, Daniels B, et al. (2017) Psychotropic medication
use in Australia, 2007 to 2015: Changes in annual incidence, preva-
lence and treatment exposure. Australian and New Zealand Journal
of Psychiatry 51: 990–999.
Brown HS, Ito K, Galetin A, et al. (2005) Prediction of in vivo drug-
drug interactions from in vitro data: Impact of incorporating parallel
pathways of drug elimination and inhibitor absorption rate constant.
British Journal of Clinical Pharmacology 60: 508–518.
Burkovski A and Krämer R (2002) Bacterial amino acid transport proteins:
Occurrence, functions, and significance for biotechnological applica-
tions. Applied Microbiology and Biotechnology 58: 265–274.
Carabotti M, Scirocco A, Maselli MA, et al. (2015) The gut-brain axis:
Interactions between enteric microbiota, central and enteric nervous
systems. Annals of Gastroenterology 28: 203.
Childs SL, Chyall LJ, Dunlap JT, et al. (2004) Crystal engineering
approach to forming cocrystals of amine hydrochlorides with organic
acids. Molecular complexes of fluoxetine hydrochloride with ben-
zoic, succinic, and fumaric acids. Journal of the American Chemical
Society 126: 13335–13342.
Coordinators NR (2018) Database resources of the national center for bio-
technology information. Nucleic Acids Research 46: D8–D13.
Coyte KZ, Schluter J and Foster KR (2015) The ecology of the micro-
biome: Networks, competition, and stability. Science 350: 663–666.
Cryan JF and Dinan TG (2019) Talking about a microbiome revolution.
Nature Microbiology 4: 552–553.
Cussotto S, Strain CR, Fouhy F, et al. (2019) Differential effects of psy-
chotropic drugs on microbiome composition and gastrointestinal
function. Psychopharmacology 236: 1671–1685.
David LA, Maurice CF, Carmody RN, et al. (2014) Diet rapidly and repro-
ducibly alters the human gut microbiome. Nature 505: 559–563.
Davies J, Spiegelman GB and Yim G (2006) The world of subinhibitory
antibiotic concentrations. Current Opinion in Microbiology 9: 445–
453.
Davis S, Hardy J and Fara J (1986) Transit of pharmaceutical dosage
forms through the small intestine. Gut 27: 886–892.
De Vane CL, Liston HL and Markowitz JS (2002) Clinical pharmacokinet-
ics of sertraline. Clinical Pharmacokinetics 41: 1247–1266.
Deak K, Takacs-Novak K, Tihanyi K, et al. (2006) Physico-chemical pro-
filing of antidepressive sertraline: Solubility, ionisation, lipophilicity.
Medicinal Chemistry 2: 385–389.
Desbonnet L, Clarke G, Traplin A, et al. (2015) Gut microbiota depletion
from early adolescence in mice: Implications for brain and behaviour.
Brain, Behavior, and Immunity 48: 165–173.
Donaldson GP, Lee SM and Mazmanian SK (2015) Gut biogeography of
the bacterial microbiota. Nature Reviews Microbiology 14: 20–32.
Eckburg PB, Bik EM, Bernstein CN, et al. (2005) Diversity of the human
intestinal microbial flora. Science 308: 1635–1638.
Edgar RC (2004) MUSCLE: Multiple sequence alignment with high accu-
racy and high throughput. Nucleic Acids Research 32: 1792–1797.
Australian & New Zealand Journal of Psychiatry, 53(12)
ANZJP Articles
Eisen JA, Dethlefsen L, Huse S, et al. (2008) The pervasive effects of
an antibiotic on the human gut microbiota, as revealed by deep 16S
rRNA Sequencing. PLoS Biology 6: e280.
Filipski KJ, Varma MV, El-Kattan AF, et al. (2013) Intestinal targeting of
drugs: Rational design approaches and challenges. Current Topics in
Medicinal Chemistry 13: 776–802.
Flowers SA, Evans SJ, Ward KM, et al. (2017) Interaction between atypi-
cal antipsychotics and the gut microbiome in a bipolar disease cohort.
Pharmacotherapy 37: 261–267.
Forslund K, Hildebrand F, Nielsen T, et al. (2015) Disentangling type 2
diabetes and metformin treatment signatures in the human gut micro-
biota. Nature 528: 262–266.
Frase L, Doerr JP, Feige B, et al. (2018) Different endocrine effects of an
evening dose of amitriptyline, escitalopram, and placebo in healthy par-
ticipants. Clinical Psychopharmacology and Neuroscience 16: 253–261.
Frohlich EE, Farzi A, Mayerhofer R, et al. (2016) Cognitive impairment
by antibiotic-induced gut dysbiosis: Analysis of gut microbiota-brain
communication. Brain, Behavior, and Immunity 56: 140–155.
Fung TC, Olson CA and Hsiao EY (2017) Interactions between the micro-
biota, immune and nervous systems in health and disease. Nature
Neuroscience 20: 145–155.
Glintborg D, Altinok ML, Ravn P, et al. (2018) Adrenal activity and meta-
bolic risk during randomized escitalopram or placebo treatment in
PCOS. Endocrine Connections 7: 479–489.
Grouleff J, Koldsø H, Miao Y, et al. (2016) Ligand binding in the extracel-
lular vestibule of the neurotransmitter transporter homologue LeuT.
ACS Chemical Neuroscience 8: 619–628.
Guida F, Turco F, Iannotta M, et al. (2018) Antibiotic-induced microbiota
perturbation causes gut endocannabinoidome changes, hippocampal
neuroglial reorganization and depression in mice. Brain, Behavior,
and Immunity 67: 230–245.
Hancock RE and Bell A (1989) Antibiotic Uptake into Gram-negative
Bacteria: Perspectives in Antiinfective Therapy. New York: Springer,
pp. 42–53.
Hendricks O (2007) Antimicrobial effects of selected non-antibiotics
on sensitivity and invasion of gram-positive bacteria. PhD Thesis,
University of Southern Denmark, Odense.
Hiemke C and Härtter S (2000) Pharmacokinetics of selective serotonin
reuptake inhibitors. Pharmacology & Therapeutics 85: 11–28.
Hoban AE, Moloney RD, Golubeva AV, et al. (2016) Behavioural and neu-
rochemical consequences of chronic gut microbiota depletion during
adulthood in the rat. Neuroscience 339: 463–477.
Huffnagle GB and Noverr MC (2013) The emerging world of the fungal
microbiome. Trends in Microbiology 21: 334–341.
Imhann F, Bonder MJ, Vila AV, et al. (2016) Proton pump inhibitors affect
the gut microbiome. Gut 65: 740–748.
Isaac S, Scher JU, Djukovic A, et al. (2017) Short- and long-term effects
of oral vancomycin on the human intestinal microbiota. Journal of
Antimicrobial Chemotherapy 72: 128–136.
Jackson MA, Goodrich JK, Maxan M-E, et al. (2016) Proton pump
inhibitors alter the composition of the gut microbiota. Gut 65:
749–756.
Jackson MA, Verdi S, Maxan ME, et al. (2018) Gut microbiota associa-
tions with common diseases and prescription medications in a popu-
lation-based cohort. Nature Communications 9: 2655.
Jeon SW and Kim YK (2016) Molecular neurobiology and promising new
treatment in depression. International Journal of Molecular Sciences
17: 381.
Jin M, Lu J, Chen Z, et al. (2018) Antidepressant fluoxetine induces
multiple antibiotics resistance in Escherichia coli via ROS-mediated
mutagenesis. Environment International 120: 421–430.
Jovanović D, Kilibarda V, Dordević S, et al. (2006) Bioequivalence testing
of a new tablet formulation of generic fluoxetine. European Journal
of Drug Metabolism and Pharmacokinetics 31: 35–40.
Kaatz GW, Moudgal VV, Seo SM, et al. (2003) Phenylpiperidine selec-
tive serotonin reuptake inhibitors interfere with multidrug efflux
McGovern et al.
pump activity in Staphylococcus aureus. International Journal of
Antimicrobial Agents 22: 254–261.
Kalaycı S, Demirci S and Sahin F (2014) Antimicrobial properties of vari-
ous psychotropic drugs against broad range microorganisms. Current
Psychopharmacology 3: 195–202.
Kasubuchi M, Hasegawa S, Hiramatsu T, et al. (2015) Dietary gut micro-
bial metabolites, short-chain fatty acids, and host metabolic regula-
tion. Nutrients 7: 2839–2849.
Keller MB, Trivedi MH, Thase ME, et al. (2007) The prevention of
recurrent episodes of depression with venlafaxine for two years
(PREVENT) study: Outcomes from the acute and continuation
phases. Biological Psychiatry 62: 1371–1379.
Kim S, Thiessen PA, Bolton EE, et al. (2016) PubChem substance and
compound databases. Nucleic Acids Research 44: D1202–1213.
Kruszewska H, Zareba T and Tyski S (2004) Examination of antimi-
crobial activity of selected non-antibiotic drugs. Acta Poloniae
Pharmaceutica 61: 18–21.
Kruszewska H, Zareba T and Tyski S (2006) Estimation of antimicro-
bial activity of selected non-antibiotic products. Acta Poloniae
Pharmaceutica 63: 457–460.
Kruszewska H, Zareba T and Tyski S (2008) Examination of antibacte-
rial and antifungal activity of selected non-antibiotic products. Acta
Poloniae Pharmaceutica: Drug Research 65: 779–782.
Kruszewska H, Zareba T and Tyski S (2010) Examination of antimi-
crobial activity of selected non-antibiotic products. Acta Poloniae
Pharmaceutica 67: 733–736.
Kumar MMK, Madhavi K, Varma PR, et al. (2015) Shape based virtual
screening and discovery of potential antitubercular activity in ‘Non-
antibiotics’. International Journal of Drug Design and Discovery
6(1): 1430–1434.
Kumar S, Stecher G and Tamura K (2016) MEGA7: Molecular evolu-
tionary genetics analysis version 7.0 for bigger datasets. Molecular
Biology and Evolution 33: 1870–1874.
Kundu P, Blacher E, Elinav E, et al. (2017) Our gut microbiome: The
evolving inner self. Cell 171: 1481–1493.
Lass-Flörl C, Dierich M, Fuchs D, et al. (2001) Antifungal properties of
selective serotonin reuptake inhibitors against Aspergillus species in
vitro. Journal of Antimicrobial Chemotherapy 48: 775–779.
Leclercq S, Mian FM, Stanisz AM, et al. (2017) Low-dose penicillin in
early life induces long-term changes in murine gut microbiota, brain
cytokines and behavior. Nature Communications 8: 15062.
Lemberger L, Bergstrom R, Wolen R, et al. (1985) Fluoxetine: Clinical
pharmacology and physiologic disposition. The Journal of Clinical
Psychiatry 46: 14–19.
Lozupone CA, Stombaugh JI, Gordon JI, et al. (2012) Diversity, stability
and resilience of the human gut microbiota. Nature 489: 220–230.
Lukic I, Getselter D, Ziv O, et al. (2019) Antidepressants affect gut micro-
biota and Ruminococcus flavefaciens is able to abolish their effects on
depressive-like behavior. Translational Psychiatry 9: 133.
Lyte M and Brown DR (2018) Evidence for PMAT- and OCT-like biogenic
amine transporters in a probiotic strain of Lactobacillus: Implications
for interkingdom communication within the microbiota-gut-brain
axis. PloS ONE 13: e0191037.
Lyte M, Daniels KM and Schmitz-Esser S (2019) Fluoxetine-induced
alteration of murine gut microbial community structure: Evidence for
a microbial endocrinology-based mechanism of action responsible for
fluoxetine-induced side effects. PeerJ 7: e6199.
Macedo D, Filho A, Soares de Sousa CN, et al. (2017) Antidepressants,
antimicrobials or both? Gut microbiota dysbiosis in depression and
possible implications of the antimicrobial effects of antidepressant
drugs for antidepressant effectiveness. Journal of Affective Disorders
208: 22–32.
Maier L, Pruteanu M, Kuhn M, et al. (2018) Extensive impact of non-
antibiotic drugs on human gut bacteria. Nature 555: 623.
Martino M, Rocchi G, Escelsior A, et al. (2012) Immunomodulation
mechanism of antidepressants: Interactions between sero-
1165
tonin/norepinephrine balance and Th1/Th2 balance. Current
Neuropharmacology 10: 97–123.
Mawe GM and Hoffman JM (2013) Serotonin signalling in the gut:
Functions, dysfunctions and therapeutic targets. Nature Reviews
Gastroenterology & Hepatology 10: 473–486.
Mehta RS, Abu-Ali GS, Drew DA, et al. (2018) Stability of the human
faecal microbiome in a cohort of adult men. Nature Microbiology 3:
347–355.
Mithani SD, Bakatselou V, TenHoor CN, et al. (1996) Estimation of the
increase in solubility of drugs as a function of bile salt concentration.
Pharmaceutical Research 13: 163–167.
Möhle L, Mattei D, Heimesaat Markus M, et al. (2016) Ly6Chi monocytes
provide a link between antibiotic-induced changes in gut microbiota
and adult hippocampal neurogenesis. Cell Reports 15: 1945–1956.
Moloney RD, Desbonnet L, Clarke G, et al. (2014) The microbiome:
Stress, health and disease. Mammalian Genome 25: 49–74.
Mould DR and Upton RN (2013) Basic concepts in population modeling,
simulation, and model-based drug development – part 2: Introduction
to pharmacokinetic modeling methods. CPT: Pharmacometrics &
Systems Pharmacology 2: e38.
Moya A and Ferrer M (2016) Functional redundancy-induced stability of
gut microbiota subjected to disturbance. Trends in Microbiology 24:
402–413.
Mrazek DA, Hornberger JC, Altar CA, et al. (2014) A review of the clini-
cal, economic, and societal burden of treatment-resistant depression:
1996–2013. Psychiatric Services 65: 977–987.
Munoz-Bellido J, Munoz-Criado S and Garcıa-Rodrıguez J (2000)
Antimicrobial activity of psychotropic drugs: Selective serotonin
reuptake inhibitors. International Journal of Antimicrobial Agents 14:
177–180.
Myhrvold C, Kotula JW, Hicks WM, et al. (2015) A distributed cell divi-
sion counter reveals growth dynamics in the gut microbiota. Nature
Communications 6: 10039.
Nikisch G, Mathe AA, Czernik A, et al. (2005) Long-term citalopram
administration reduces responsiveness of HPA axis in patients with
major depression: Relationship with S-citalopram concentrations
in plasma and cerebrospinal fluid (CSF) and clinical response.
Psychopharmacology 181: 751–760.
O’Leary OF, Ogbonnaya ES, Felice D, et al. (2018) The vagus nerve
modulates BDNF expression and neurogenesis in the hippocampus.
European Neuropsychopharmacology 28: 307–316.
O’Mahony SM, Clarke G, Borre YE, et al. (2015) Serotonin, tryptophan
metabolism and the brain-gut-microbiome axis. Behavioural Brain
Research 277: 32–48.
Opperman TJ and Nguyen ST (2015) Recent advances toward a molecular
mechanism of efflux pump inhibition. Frontiers in Microbiology 6: 421.
Ortega-Martinez S (2015) A new perspective on the role of the CREB
family of transcription factors in memory consolidation via adult hip-
pocampal neurogenesis. Frontiers in Molecular Neuroscience 8: 46.
Patel JB (2017) Performance Standards for Antimicrobial Susceptibility
Testing. Wayne, PA: Clinical and Laboratory Standards Institute.
Pawar HA, Ayre AP and Lalitha K (2012) Development of validated ana-
lytical method for in-vitro dissolution study of sertraline hydrochlo-
ride capsules. Journal of Current Pharma Research 2: 560–565.
Powell N, Walker MM and Talley NJ (2017) The mucosal immune sys-
tem: Master regulator of bidirectional gut-brain communications.
Nature Reviews Gastroenterology & Hepatology 14: 143–159.
Pratt LA, Brody DJ and Gu Q (2017) Antidepressant Use among Persons
Aged 12 and Over: United States, 2011-2014. National Center for
Health Statistics Data Brief 283: 1–8.
Qamar F and Safila N (2015) Comparrative assay of citalopram in differ-
ent media. Journal of Bioequivalence & Bioavailability 7: 4.
Ramsteijn AS, Jašarević E, Houwing DJ, et al. (2018) Antidepressant
treatment modulates the gut microbiome and metabolome during
pregnancy and lactation in rats with a depressive-like phenotype.
Epub ahead of print 19 December. DOI: 10.1101/501742.
Australian & New Zealand Journal of Psychiatry, 53(12)
1166
Rhee K-J, Sethupathi P, Driks A, et al. (2004) Role of commensal bacteria
in development of gut-associated lymphoid tissues and preimmune
antibody repertoire. The Journal of Immunology 172: 1118–1124.
Rogers MAM and Aronoff DM (2016) The influence of non-steroidal anti-
inflammatory drugs on the gut microbiome. Clinical Microbiology
and Infection 22: 178. e171–e179.
Rojo D, Méndez-García C, Raczkowska BA, et al. (2017) Exploring the
human microbiome from multiple perspectives: Factors altering its
composition and function. FEMS Microbiology Reviews 41: 453–478.
Rooks MG and Garrett WS (2016) Gut microbiota, metabolites and host
immunity. Nature Reviews Immunology 16: 341–352.
Rush AJ, Trivedi MH, Wisniewski SR, et al. (2006) Acute and longer-term out-
comes in depressed outpatients requiring one or several treatment steps:
A STAR* D report. American Journal of Psychiatry 163: 1905–1917.
Sahay A and Hen R (2007) Adult hippocampal neurogenesis in depression.
Nature Neuroscience 10: 1110.
Schemann M and Neunlist M (2004) The human enteric nervous system.
Neurogastroenterology & Motility 16: 55–59.
Schiller C, Frohlich CP, Giessmann T, et al. (2005) Intestinal fluid vol-
umes and transit of dosage forms as assessed by magnetic resonance
imaging. Alimentary Pharmacology & Therapeutics 22: 971–979.
Seaman JS, Bowers SP, Dixon P, et al. (2005) Dissolution of common
psychiatric medications in a Roux-en-Y gastric bypass model.
Psychosomatics 46: 250–253.
Shameem M, Katori N, Aoyagi N, et al. (1995) Oral solid controlled
release dosage forms: Role of GI-mechanical destructive forces and
colonic release in drug absorption under fasted and fed conditions in
humans. Pharmaceutical Research 12: 1049–1054.
Sim K, Lau WK, Sim J, et al. (2015) Prevention of relapse and recur-
rence in adults with major depressive disorder: Systematic review
and meta-analyses of controlled trials. The International Journal of
Neuropsychopharmacology 19: pyv076.
Singh SK (2014) LeuT: A prokaryotic stepping stone on the way to a eukar-
yotic neurotransmitter transporter structure. Channels 2: 380–389.
Singh SK, Yamashita A and Gouaux E (2007) Antidepressant binding site
in a bacterial homologue of neurotransmitter transporters. Nature 448:
952–956.
Stilling RM, van de Wouw M, Clarke G, et al. (2016) The neuropharma-
cology of butyrate: The bread and butter of the microbiota-gut-brain
axis? Neurochemistry International 99: 110–132.
Sutton SC (2009) Role of physiological intestinal water in oral absorption.
The AAPS Journal 11: 277–285.
Australian & New Zealand Journal of Psychiatry, 53(12)
ANZJP Articles
Ticinesi A, Milani C, Lauretani F, et al. (2017) Gut microbiota composi-
tion is associated with polypharmacy in elderly hospitalized patients.
Scientific Reports 7: 11102.
van Gorp F, Duffull S, Hackett LP, et al. (2012) Population pharmacoki-
netics and pharmacodynamics of escitalopram in overdose and the
effect of activated charcoal. British Journal of Clinical Pharmacology
73: 402–410.
van Harten J (1995) Overview of the pharmacokinetics of fluvoxamine.
Clinical Pharmacokinetics 29: 1–9.
Venkatakrishnan K and Obach RS (2005) In vitro-in vivo extrapolation
of CYP2D6 inactivation by paroxetine: Prediction of nonstationary
pharmacokinetics and drug interaction magnitude. Drug Metabolism
and Disposition 33: 845–852.
Vogt NM, Kerby RL, Dill-McFarland KA, et al. (2017) Gut microbiome
alterations in Alzheimer’s disease. Scientific Reports 7: 13537.
Wade AG, Crawford GM and Yellowlees A (2011) Efficacy, safety and
tolerability of escitalopram in doses up to 50 mg in major depres-
sive disorder (MDD): An open-label, pilot study. BMC Psychiatry
11: 42.
Wernert GT, Winkler DA, Holan G, et al. (2004) Synthesis, biologi-
cal activity, and QSAR studies of antimicrobial agents containing
biguanide isosteres. Australian Journal of Chemistry 57: 77–85.
Winter G, Hart RA, Charlesworth RPG, et al. (2018) Gut microbiome
and depression: What we know and what we need to know. Nature
Reviews Neuroscience 29: 629–643.
Yamashita A, Singh SK, Kawate T, et al. (2005) Crystal structure of a bac-
terial homologue of Na+/Cl--dependent neurotransmitter transport-
ers. Nature 437: 215–223.
Yano Jessica M, Yu K, Donaldson Gregory P, et al. (2015) Indigenous
bacteria from the gut microbiota regulate host serotonin biosynthesis.
Cell 161: 264–276.
Yee AL and Gilbert JA (2016) Is triclosan harming your microbiome?
Science 353: 348–349.
Yim G, Wang HH and Davies J (2007) Antibiotics as signalling mole-
cules. Philosophical Transactions of the Royal Society of London B:
Biological Sciences 362: 1195–1200.
Zheng P, Zeng B, Zhou C, et al. (2016) Gut microbiome remodeling
induces depressive-like behaviors through a pathway mediated by the
host’s metabolism. Molecular Psychiatry 21: 786–796.
Zimmermann M, Zimmermann-Kogadeeva M, Wegmann R, et al. (2019)
Mapping human microbiome drug metabolism by gut bacteria and
their genes. Nature 570: 462–467.