BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from

BMJ Open Science is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.

When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.

The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.

BMJ Open Science is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay-per-view fees (http://openscience.bmj.com). ​ ​

If you have any questions on BMJ Open’s open peer review process please email [email protected] ​

http://openscience.bmj.com/ on September 27, 2021 by guest. Protected copyright. BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from

BMJ Open Science

Confidential: For Review Only A protocol for systematic review and meta-analysis of the evidence linking hippocampal neurogenesis to the effects of on mood and behaviour

Journal: BMJ Open Science

Manuscript ID bmjos-2020-100077

Article Type: Protocol

Date Submitted by the 02-Apr-2020 Author:

Complete List of Authors: Bolzan, Juliana; Universidade Federal de Santa Catarina, Departamento de Ciências Fisiológica; Universidade Federal de Santa Catarina, Pós- Graduação em Farmacologia Lino de Oliveira, Cilene; Universidade Federal de Santa Catarina, Departamento de Ciências Fisiológicas; Universidade Federal de Santa Catarina, Pós-Graduação em Farmacologia

http://openscience.bmj.com/ on September 27, 2021 by guest. Protected copyright.

https://mc.manuscriptcentral.com/bmjos Page 1 of 21 BMJ Open Science BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 4 5 6 7 8 9 I, the Submitting Author has the right to grant and does grant on behalf of all authors of the Work (as defined 10 in the below author licence), an exclusive licence and/or a non-exclusive licence for contributions from authors 11 who are: i) UK Crown employees; ii) where BMJ has agreed a CC-BY licence shall apply, and/or iii) in accordance 12 with the termsConfidential: applicable for US Federal Government For officers Review or employees actingOnly as part of their official 13 duties; on a worldwide, perpetual, irrevocable, royalty-free basis to BMJ Publishing Group Ltd (“BMJ”) its 14 licensees and where the relevant Journal is co-owned by BMJ to the co-owners of the Journal, to publish the 15 Work in this journal and any other BMJ products and to exploit all rights, as set out in our licence. 16 17 The Submitting Author accepts and understands that any supply made under these terms is made by BMJ to 18 the Submitting Author unless you are acting as an employee on behalf of your employer or a postgraduate 19 student of an affiliated institution which is paying any applicable article publishing charge (“APC”) for Open 20 Access articles. Where the Submitting Author wishes to make the Work available on an Open Access basis (and 21 intends to pay the relevant APC), the terms of reuse of such Open Access shall be governed by a Creative 22 Commons licence – details of these licences and which Creative Commons licence will apply to this Work are set 23 out in our licence referred to above. 24 25 Other than as permitted in any relevant BMJ Author’s Self Archiving Policies, I confirm this Work has not been 26 accepted for publication elsewhere, is not being considered for publication elsewhere and does not duplicate 27 material already published. I confirm all authors consent to publication of this Work and authorise the granting 28 of this licence. 29 30 31 32 33 34 35 http://openscience.bmj.com/ 36 37 38 39 40 41 42 43

44 on September 27, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 2 of 21

1 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 Title: A protocol for systematic review and meta-analysis of the evidence linking 4 5 6 hippocampal neurogenesis to the effects of antidepressants on mood and behaviour 7 8 Authors Juliana A Bolzan 1,2, Cilene Lino de Oliveira1,2. 9 10 11 Affiliations: 12 Confidential: For Review Only 13 14 1- Department of Physiological Sciences, Center of Biological Sciences, Federal 15 16 University of Santa Catarina – UFSC, Campus Trindade, 88040-900, Florianópolis - SC, 17 18 Brazil. 19 20 21 2- Post graduation in Pharmacology, Center of Biological Sciences, Federal University 22 23 24 of Santa Catarina – UFSC, Campus Trindade, 88040-900, Florianópolis - SC, Brazil. 25 26 Address to correspondence: 27 28 29 Cilene Lino de Oliveira, LABORATORY OF BEHAVIORAL NEUROBIOLOGY- 30 31 Department of Physiological Sciences, Center of Biological Sciences, Federal 32 33 34 University of Santa Catarina, University Campus, Trindade, CEP: 88040-900, 35 http://openscience.bmj.com/ 36 Florianópolis, SC – Brazil, Phone: +55 48 3721-4689, E-mail: [email protected], 37 38 Orcid: orcid.org/0000-0002-0627-530X 39 40 41 Running title: Hippocampal neurogenesis and antidepressants. 42 43 Word counting (except title page, abstract, references, figures, tables or 44 on September 27, 2021 by guest. Protected copyright. 45 46 acknowledgments): 2,174 47 48 49 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos Page 3 of 21 BMJ Open Science

2 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 Authors’ contributions: 4 5 6 CLO: selected research theme; performed preliminary studies; wrote the preliminary 7 8 protocol; wrote the manuscript; approved final version of the manuscript. 9 10 JAB: performed preliminary studies; wrote the preliminary protocol; wrote the 11 12 manuscript;Confidential: approved the final version For of the manuscript. Review Only 13 14 15 Funding: This study was financed in part by the Coordenação de Aperfeiçoamento de 16 17 Pessoal de Nível Superior – Brasil (CAPES) – Finance Code 001. Juliana A Bolzan 18 19 receive a fellowship from Conselho Nacional de Desenvolvimento Científico e 20 21 22 Tecnológico, Brazil (131247/2019-0). 23 24 Competing interests: Senior author (Cilene Lino de Oliveira) is an Associate Editor of 25 26 BMJ Open Science. Other than that, no conflict of interest is declared. 27 28 29 Open data: open data are available at (https://osf.io/5z8us/) 30 31 32 33 34 35 http://openscience.bmj.com/ 36 37 38 39 40 41 42 43

44 on September 27, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 4 of 21

3 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 ABSTRACT 4 5 6 Objective: Rodent studies associated deficits of hippocampal neurogenesis with 7 8 depression and the reversal of these deficits to effects. A previous 9 10 systematic review (SR) and meta-analysis (MA) suggested that pro neurogenic effects of 11 12 antidepressantsConfidential: may vary within compounds. For TheReview present protocol Only aims to evaluate the 13 14 15 evidence linking hippocampal neurogenesis to the effects of antidepressants on 16 17 behaviour. Search strategy, Screening annotation, Data management: Searches in 18 19 Medline (via PubMed) and Web of Science will provide publications reporting studies 20 21 22 effects of antidepressants on neurogenesis in the hippocampus of adult rodents without 23 24 (SR 1) or with behavioural testing (SR 2). The SR 3 aim to obtain studies reporting effects 25 26 of ablation of hippocampal neurogenesis on the actions of antidepressants in animal 27 28 models. Application of a broad list of inclusion criteria will exclude studies reporting co- 29 30 31 treatment data and reviews. Quality assessment of the included publications will use the 32 33 CAMARADES checklist and Syrcles’RoB tool. If feasible, the random-effect model MA 34 http://openscience.bmj.com/ 35 will analyze the direction, magnitude and significance of effect sizes within subgroups 36 37 38 (meta-regression, and if necessary, stratified regression). Heterogeneity will be calculated 39 40 using chi-square statistics with n-1 degrees of freedom. Funnel plotting, Egger regression, 41 42 'trim and fill' will be used to estimate the risk of publication bias. Due to the continuous 43

44 on September 27, 2021 by guest. Protected copyright. 45 production in the field of hippocampal neurogenesis implementation of a “Living 46 47 Systematic Review” is intended. Reporting: Find a preliminary version of this protocol 48 49 at the Open Science Framework (https://osf.io/gmsvd/). The results will be published in 50 51 a peer-reviewed scientific journal. 52 53 54 Key words: hippocampus, neurogenesis, antidepressants, effect size, heterogeneity, 55 56 animal models. 57 58 59 60

https://mc.manuscriptcentral.com/bmjos Page 5 of 21 BMJ Open Science

4 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 STRENGTHS AND LIMITATIONS OF THIS STUDY 4 5  6 A preliminary version of this protocol is at Open Science Framework. 7 8  A broad list of inclusion criteria will allow for subgroup analysis. 9 10  Primary studies often compare several subgroups bringing possible correlations 11 12 betweenConfidential: independent variables. For Review Only 13 14 15  High heterogeneity may limit the external validity of the results. 16 17  Publication bias may inflate the estimated effect size. 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 http://openscience.bmj.com/ 36 37 38 39 40 41 42 43

44 on September 27, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 6 of 21

5 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 INTRODUCTION 4 5 6 The consumption of antidepressant used in psychiatric treatments 7 8 doubled according to the organization of Economic Cooperation and Development 9 10 (OECD) between the years 2000 and 2015 [1]. Despite a large number of patients under 11 12 treatment,Confidential: the mechanisms underlying For therapeutic Review effects of antidepressantsOnly remain 13 14 15 elusive. Behavioural studies indicate that the therapeutic effects of antidepressants 16 17 depend on the reversal of neurochemical deficiencies [2]. For example, earlier studies 18 19 revealed that most antidepressants inhibit the reuptake of monoamines promoting high 20 21 22 extracellular concentrations of , noradrenaline, in the central nervous 23 24 system (CNS) [3]. 25 26 Moreover, antagonism of transmission provoked sudden decay of 27 28 the effectiveness of antidepressants in patients [4, 5]. Although durable, the relationships 29 30 31 between monoamines, depression and antidepressants may not explain all aspects of 32 33 antidepressant therapy. For example, the delay between the onset of treatment and 34 http://openscience.bmj.com/ 35 remission of symptoms or the existence of depressions resistant to prototype 36 37 38 antidepressants suggests there are more for depression than monoamines. 39 40 Animal models allow for discoveries generating hypothesis and theories on 41 42 depression and antidepressant efficacy beyond monoamines. Neurogenic theory of 43

44 on September 27, 2021 by guest. Protected copyright. 45 depression gained more attention in the 2000s when a series of papers reported that 46 47 chronic antidepressant treatment increased the number of newborn neurons in the 48 49 hippocampus of adult rodents [6,7,8]. Stressful stimuli provoked hippocampal cell loss 50 51 prevented by chronic antidepressant treatment in rats [9]. Furthermore, ablation of 52 53 54 hippocampal neurogenesis reverted the effects of antidepressants on the behaviour of 55 56 adult mice [7,8,10]. Altogether, these data engendered an explanation considering 57 58 hippocampal neurogenesis as the underlying mechanism for the effects of antidepressants 59 60

https://mc.manuscriptcentral.com/bmjos Page 7 of 21 BMJ Open Science

6 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 on mood and behaviour. Since the neurogenesis is a time-consuming process, the 4 5 6 neurogenic theory offers a plausible explanation to delay between the onset of 7 8 antidepressant treatment and the therapeutic effect. Neurogenic theory inspired a long 9 10 series of studies aiming to understand details of the process which could provide a variety 11 12 of newConfidential: targets for the development of Forantidepressants. Review Only 13 14 15 Over time the findings diverging of the earliest publications emerged indicating 16 17 that some effects of antidepressants could be related to hippocampal neurogenesis while 18 19 others would be unrelated to it [11,12]. For example, the effects of chronic antidepressants 20 21 22 on behavioural tests such as novelty suppressed-feeding seem to correlate positively and 23 24 depend on hippocampal neurogenesis [7]. In contrast, the effects of antidepressants on 25 26 the forced swimming test seem unrelated to it [11]. A systematic review (SR) and meta- 27 28 analysis (MA) of studies investigating effects of antidepressants on the scores of 29 30 31 neurogenic markers in the hippocampus of adult laboratory rodents suggested that some 32 33 antidepressants increased the outcomes while others did not [13]. Depending on the type 34 http://openscience.bmj.com/ 35 of the compound (fluoxetine, , , ), laboratory 36 37 38 species (rat or mice) and the type of molecular marker (BrdU, doublecortin, Ki-67), the 39 40 magnitude of the effect size (ES) may vary from negligible to very large [13]. The broad 41 42 range of ES values may indicate that actions of antidepressants may be more or less 43

44 on September 27, 2021 by guest. Protected copyright. 45 related or dependent on hippocampal neurogenesis. Although pioneer, the SRMA 46 47 mentioned above, analysed studies published on a recent period comprising five years 48 49 (2013-2018) which may lessen the power of the analysis and reliability of the conclusions. 50 51 This protocol presents methods to perform a more comprehensive SR and MA aiming to 52 53 54 investigate the strength of the functional relationships between antidepressant treatments 55 56 and hippocampal neurogenesis. 57 58 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 8 of 21

7 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 For convenience, present work classifies relationships between hippocampal 4 5 6 neurogenesis and antidepressants as correlational or functional. Studies reporting 7 8 correlations often evaluate hippocampal neurogenesis post-mortem in animals treated 9 10 with antidepressants in vivo [14]. Sometimes correlational studies also evaluate the 11 12 behaviouralConfidential: effects of antidepressants For in vivoReview [15]. Functional Only studies are those 13 14 15 manipulating the levels of hippocampal neurogenesis to evaluate its relevance to the 16 17 effects of antidepressants in behavioural tests, for example, see [7]. Due to the variety of 18 19 primary studies investigating the theme, it seems not feasible to obtain all relevant 20 21 22 publications with a single SR. Therefore, the planning of three different SR was guided 23 24 by the following questions: 1- do different antidepressants increase hippocampal 25 26 neurogenesis at different degrees? 2- are there correlations between behavioral and 27 28 neurogenic effects of antidepressants in animal models? 3- is there a causal link between 29 30 31 behavioural and neurogenic effects of antidepressants in animal models? 32 33 Broad inclusion and exclusion criteria will screen correlational studies for the SR 34 http://openscience.bmj.com/ 35 1 and 2 and functional studies for the SR 3. In the first round of screening, examining 36 37 38 titles and abstracts, the reviews, systematic reviews and meta-analysis will be excluded. 39 40 Examining the full texts will allow for the application of inclusion criteria related to 41 42 population (rat and mice of any sex, age or strain), interventions (antidepressants or 43

44 on September 27, 2021 by guest. Protected copyright. 45 ablation of neurogenesis), control (vehicle) or outcome (molecular markers of 46 47 neurogenesis or behaviours). A preliminary search in the Medline, retrieved more than 48 49 1,249 publications for the SR 1 and 145 publications for the SR 2. After screening, SR 1 50 51 included 66 references, and SR 2 included 51 ones. No preliminary study occurred for 52 53 54 SR 3. At the end of the screening process, parameters will be extracted to estimate: (1) 55 56 quality; (2) ES; (3) Heterogeneity; and (4) Publication bias. The “Systematic Review 57 58 Protocol for Animal Interventions Studies” (SYRCLE) [16] is the template to this protocol, 59 60

https://mc.manuscriptcentral.com/bmjos Page 9 of 21 BMJ Open Science

8 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 which has an early version at the Open Science Framework (OSF, June 18, 2019, 4 5 6 https://osf.io/gmsvd/). The quality of publications will be evaluated by the use of the 7 8 checklist adapted from the Collaborative Approach to Meta-Analysis and Review of 9 10 Animal Data from Experimental Studies (CAMARADES) [17] and Risk of Bias of Syrcle 11 12 (SyrclesConfidential: RoB tool) [18]. For Review Only 13 14 15 METHODS AND ANALYSIS 16 17 Overview of the systematic review 18 19 The plan of every systematic review (SR 1, SR 2, SR 3) included the following 20 21 22 steps: 1-research question; 2-choice of search strategy; 3-choice of bibliographic 23 24 databases; 4- choice of inclusion and exclusion criteria; 5- planning of data extraction and 25 26 6- planning of the meta-analysis (Figure 1). SR 1, 2 and 3 aim to answer the following 27 28 questions prepared using the PICO tool [19], respectively: 1-do different antidepressants 29 30 31 increase hippocampal neurogenesis at different degrees? 2-are there correlations between 32 33 behavioural and neurogenic effects of antidepressants in animal models? 3- is there a 34 http://openscience.bmj.com/ 35 causal link between behavioural and neurogenic effects of antidepressants in animal 36 37 38 models? The search strategy and combination of keywords are from protocols previously 39 40 published [20]. Search in Medline will be conducted using the platform PubMed (advanced 41 42 search, http://www.ncbi.nlm.nih) and the Web of Science with the platform “Periódicos 43

44 on September 27, 2021 by guest. Protected copyright. 45 Capes” (https://www.periodicos.capes.gov.br/). 46 47 Search strategy 48 49 The keywords for SR 1 are a combination of MeSH terms related to the primary 50 51 outcome and intervention: ‘neurogenesis’ and ‘hippocampus’ and (or ‘BrdU’ or ‘DCX’ 52 53 54 or ‘Ki-67’, #1-5) and ‘antidepressants’ (#6) (Table 1). 55 56 Table 1 - Keywords for the search in the Medline (via Pubmed) database [20]. 57 58 “neurogenesis” [MeSH] #1 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 10 of 21

9 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 “Hippocampus” [MeSH] #2 4 5 6 “BrdU” [MeSH] #3 7 8 “Ki-67” [MeSH] #4 9 10 “DCX” [MeSH] #5 11 12 Confidential: For Review Only ((((((((ssri) OR (ssris) OR (selective serotonin reuptake inhibitor) OR (selective #6 13 14 15 serotonin reuptake inhibitors) OR (selective serotonin re-uptake inhibitor) OR (selective 16 17 serotonin re-uptake inhibitors) OR (selective serotonin reuptake inhibitors) OR (selective 18 19 serotonin-reuptake inhibitor) OR () OR () OR () OR 20 21 22 () OR () OR () OR () OR (snri) OR (ssris) 23 24 OR (serotonin and noradrenaline reuptake inhibitors) OR (serotonin and noradrenaline 25 26 reuptake inhibitor) OR (serotonin and noradrenaline re-uptake inhibitors) OR (serotonin 27 28 29 and noradrenaline re-uptake inhibitor) OR (serotonin-noradrenaline reuptake inhibitors) 30 31 OR (serotonin and norepinephrine reuptake inhibitors) OR (serotonin and 32 33 norepinephrine reuptake inhibitor) OR (serotonin and norepinephrine re-uptake 34 http://openscience.bmj.com/ 35 inhibitors) OR (serotonin and norepinephrine re-uptake inhibitor) OR (serotonin- 36 37 38 norepinephrine reuptake inhibitors) OR () OR () OR 39 40 () OR () OR (venlafaxine) OR (desvenlafaxine) OR () 41 42 OR () OR (TCA) OR (tcas) OR () OR (tricyclic 43

44 on September 27, 2021 by guest. Protected copyright. 45 antidepressants) OR (tricyclic anti-depressant) OR (tricyclic antidepressants) OR 46 47 () OR () OR () OR () OR () 48 49 OR () OR (imipramine) OR () OR () OR (melitracene) OR 50 51 52 () OR () OR () OR () OR (sari) OR 53 54 (saris) OR (serotonin antagonist and reuptake inhibitor) OR (serotonin antagonist and 55 56 reuptake inhibitors) OR (serotonin antagonist and re-uptake inhibitor) OR (serotonin 57 58 antagonist and reuptake inhibitors) OR () OR () OR 59 60

https://mc.manuscriptcentral.com/bmjos Page 11 of 21 BMJ Open Science

10 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 () OR (lubazodone) OR () OR () OR (NRI) OR (nris) 4 5 6 OR (noradrenaline reuptake inhibitor) OR (noradrenaline reuptake inhibitors) OR 7 8 (noradrenaline re-uptake inhibitor) OR (noradrenaline re-uptake inhibitors) OR 9 10 (norepinephrine reuptake inhibitor) OR (norepinephrine reuptake inhibitors) OR 11 12 Confidential: For Review Only (norepinephrine re-uptake inhibitor) OR (norepinephrine re-uptake inhibitors) OR 13 14 15 () OR () OR () OR (ndri) OR (ndri) OR (norepinephrine 16 17 dopamine reuptake inhibitor) OR (norepinephrine dopamine reuptake inhibitors) OR 18 19 (norepinephrine dopamine reuptake inhibitor) OR (norepinephrine dopamine reuptake 20 21 22 inhibitors) OR (norepinephrine dopamine reuptake inhibitor) OR (norepinephrine 23 24 dopamine reuptake inhibitors) OR (norepinephrine dopamine reuptake inhibitor) OR 25 26 (norepinephrine dopamine reuptake inhibitors) OR (norepinephrine and dopamine 27 28 29 reuptake inhibitor) OR (norepinephrine and dopamine reuptake inhibitors) OR 30 31 (norepinephrine and dopamine re-uptake inhibitor) OR (norepinephrine and dopamine 32 33 reuptake inhibitors) OR (noradrenaline dopamine reuptake inhibitor) OR (noradrenaline 34 http://openscience.bmj.com/ 35 dopamine reuptake inhibitors) OR (noradrenaline dopamine re-uptake inhibitor) OR 36 37 38 (noradrenaline dopamine re-uptake inhibitors) OR (noradrenaline dopamine reuptake 39 40 inhibitor) OR (noradrenaline dopamine reuptake inhibitors) OR (noradrenaline 41 42 dopamine re uptake inhibitor) OR (noradrenaline dopamine reuptake inhibitors) OR 43

44 on September 27, 2021 by guest. Protected copyright. 45 (noradrenaline and dopamine reuptake inhibitor) OR (noradrenaline and dopamine 46 47 reuptake inhibitors) OR (noradrenaline and dopamine re-uptake inhibitor) OR 48 49 (noradrenaline and dopamine re-uptake inhibitors) OR () OR 50 51 52 () OR () OR (ndra) OR (nras) OR (norepinephrine 53 54 dopamine releasing agent) OR (norepinephrine dopamine releasing agents) OR 55 56 (norepinephrine-dopamine releasing agent) OR (norepinephrine-dopamine releasing 57 58 agents) OR (norepinephrine and dopamine releasing agent) OR (norepinephrine and 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 12 of 21

11 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 dopamine releasing agents) OR (noradrenaline dopamine releasing agent) OR 4 5 6 (noradrenaline dopamine releasing agents) OR (noradrenaline dopamine releasing agent) 7 8 OR (noradrenaline-dopamine releasing agents) OR (noradrenaline and dopamine 9 10 releasing agent) OR (noradrenaline and dopamine releasing agents) OR () 11 12 Confidential: For Review Only OR (dextroamphetamine) OR (dextromethamphetamine) OR (lisdexamfetamine) OR 13 14 15 (teca) OR (texas) OR (tetracyclic antidepressant) OR (tetracyclic antidepressants) OR 16 17 (tetracyclic antidepressant) OR (tetracyclic anti-depressants) OR () OR 18 19 () OR () OR () OR (maoi) OR (maois) OR (monoamine 20 21 22 oxidase inhibitor) OR (monoamine oxidase inhibitors) OR (isocarboxazid) OR 23 24 (moclobemide) OR (phenelzine) OR (pirlindole) OR (selegiline) OR (tranylcypromine) 25 26 OR (antidepressant) OR (antidepressants) OR (antidepressant) OR (anti- 27 28 29 depressants)))))) 30 31 The keywords for the SR 2 are a combination of MeSH terms related to the 32 33 secondary outcomes (neurogenesis, #1-5, Table 1) and intervention (antidepressants, #6, 34 http://openscience.bmj.com/ 35 Table 1) with the primary outcomes (#7, Table 2), which were chosen by examining the 36 37 38 references included in the SR 1. 39 40 41 42 43

44 on September 27, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos Page 13 of 21 BMJ Open Science

12 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 Table 2 - Keywords for search in the Medline (via Pubmed) database. 4 5 6 (forced swimming test OR forced swimming tests OR forced swimming test, fst OR fst #7 7 8 OR forced swim OR Porsolt test OR Porsolt tests OR “fear conditioning” OR 9 10 “learning” OR “consolidation” OR “acquisition” OR “elevated plus maze” OR “plus- 11 12 Confidential: For Review Only maze” OR “open field” OR “OFT” OR “light dark” OR "Novelty suppressed feeding" 13 14 15 OR "Tail suspension test" OR "Morris water maze test" OR "Splash test" OR “water 16 17 maze test” OR “ mater maze” OR "Novelty-induced Hypophagia" OR “Isolation 18 19 rearing” OR “Novel location recognition test” OR “Rotarod Performance test” OR 20 21 22 “resident-intruder” OR “Sucrose preference test” OR “Barnes maze test”) 23 24 The keywords for the SR 3 are a combination of MeSH terms related to the 25 26 primary (#7, Table 2) and secondary outcomes (neurogenesis, #1-5, Table 1) and 27 28 29 intervention (#8, Table 3), which were chosen by examining the references similar to 30 31 Santarelli et al., 2003 [7]. 32 33 Table 3 - Keywords for search in the Medline (via Pubmed) database 34 35 http://openscience.bmj.com/ 36 ("ablation" OR "x-ray" OR "x-rays" OR "x-irradiation" OR "hippocampal irradiation" #8 37 38 OR "irradiation" OR "radiation" OR "stimulation" OR "inhibition" OR "radiological 39 40 methods" OR "radiological" OR "genetic" OR "electroconvulsive therapy" OR "ECS") 41 42 Publications returned from every search will be exported to the respective 43

44 on September 27, 2021 by guest. Protected copyright. 45 reference manager file of EndNote X7. A single investigator will evaluate the titles and 46 47 abstracts obtained to assess if they meet the broad inclusion criteria. A second investigator 48 49 will check the results of the first investigator. Discussion with a third investigator will 50 51 52 solve discrepancies. 53 54 Inclusion and exclusion criteria 55 56 Studies will be included in the SRs regardless of randomization or blinding or 57 58 59 study design (Figure 2). In these studies, interventions (antidepressants in SR 1 and 2; 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 14 of 21

13 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 ablation of neurogenesis SR 3) will be compared to controls, i.e., vehicle or the absence 4 5 6 of the intervention. Publications since 2003 (year of the seminal paper Santarelli et al., 7 8 2003 [7]) until the day of the search in databases are eligible. There will have no restriction 9 10 on language. The first round of analysis will exclude reviews, systematic reviews, meta- 11 12 analysisConfidential: and duplicates. The second round For of analysis Review will exclude Only experiments with co- 13 14 15 treatments. Studies on adult rats or mice, regardless of sex, strain or stress background; 16 17 studies with any route, dose and treatment schedule for drug administration; studies 18 19 measuring molecular markers BrdU, Ki-67 or DCX in the dentate gyrus of the 20 21 22 hippocampus are eligible. 23 24 Quality of study 25 26 Quality of every publication will be estimated by the use of CAMARADES' study 27 28 quality checklist [17], adapted with 20 points scale as follows: 1-report of the species, 29 30 31 lineage or other identifying characteristics of the experimental animal in the title, abstract 32 33 or full text; 2-peer-reviewed publication; 3-description of the method of measuring 34 http://openscience.bmj.com/ 35 behaviors; 4- report of the age, weight or stage of the experimental animal in the title, 36 37 38 abstract or full text; 5- experimental animal report in the title, abstract or full text; 6- 39 40 description of experimental animal breeding conditions in the title, abstract or full text; 41 42 7- use of experimental animals with phenotypes of interest (stressed? Depressed? 43

44 on September 27, 2021 by guest. Protected copyright. 45 Anxious?); 8- description of the accommodation conditions of the experimental animal 46 47 in the title, abstract or full text; 9-description of actions taken in relation to animal welfare 48 49 (for example: environmental enrichment); 10 - declaration of compliance with animal 50 51 testing regulations and legislation; 11- report of the exact number of animals per 52 53 54 experimental group, mean, error or standard deviation; 12- description of sample size 55 56 calculations; 13- description of antibodies and markers used for knowledge cells; 14 - 57 58 description of the method of detecting the cellular markers; 15- description of the cellular 59 60

https://mc.manuscriptcentral.com/bmjos Page 15 of 21 BMJ Open Science

14 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 quantification method; 16 - mention of the blinding (or not) of the results measures; 17 - 4 5 6 mention of treatment blinding (or not); 18- mention the randomization (or not) of the 7 8 allocation of animals to the experimental groups; 19- mentioning the exclusion (or not) 9 10 of analysis data; 20- declaration of interest. The quality of every study will also be 11 12 evaluatedConfidential: using Syrcle's Rob tool, byFor answering Review ten signalling Only questions with "yes", 13 14 [18] 15 "unclear" or "no" is considered a risk of bias "low", "unclear" and "high" respectively . 16 17 Outcome extraction 18 19 For a qualitative appraisal of the literature and subgroup analysis the following 20 21 22 information will be extracted from every publication: 1-bibliographic information 23 24 (authorship, year, journal); 2-experimental design (randomization; blinding; statistical 25 26 analysis); 3-experimental animals (species, strain, sex, age, the timing of euthanasia); 3- 27 28 antidepressants (type, dose, number of injections per day); 4-molecular markers (type, the 29 30 31 protocols for detection, protocols for BrdU injection, survival or proliferation). For 32 33 quantitative analysis, primary and secondary outcomes will be extracted. Molecular 34 http://openscience.bmj.com/ 35 markers of neurogenesis in the adult hippocampus (BrdU, Ki-67, DCX) are chosen 36 37 [13] 38 according to previously published SRMA , which will be the primary outcome in the 39 40 SR 1 and SR 2 and the secondary in SR 3. Behavioural measures will be the primary 41 42 outcome in SR 2 and 3. Estimations of cell proliferation will occur by scoring BrdU or 43

44 on September 27, 2021 by guest. Protected copyright. 45 Ki-67. Quantifying new neurons will occur by scoring co-labelling BrdU-NeuN (or BrdU 46 47 and another marker of mature neurons). The number of immature neurons will come from 48 49 the scores of DCX. Measurements may be using stereological or non-stereological 50 51 methods. Behavioural outcomes extracted will depend on the type of behavioural test 52 53 54 (e.g., time of immobility in forced swimming or tail-hanging, latency to eat in feed 55 56 suppressed by novelty). For every quantitative outcome will be extracted the number of 57 58 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 16 of 21

15 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 animals (N), mean (M), standard deviation (SDM) or standard error of the mean (SEM) 4 5 6 of control and experimental groups. 7 8 In cases of the absence of data required for meta-analysis in the text or tables of 9 10 publications, a digital rule software will be used to measurement of the outcomes in the 11 12 charts.Confidential: If information is completely unavailable For Reviewin the publication, Onlyit may request directly 13 14 15 by contact with the authors. The impossibility of obtaining the necessary data will exclude 16 17 the study from the analysis. Excel spreadsheet will receive all data extracted from the 18 19 relevant studies. 20 21 22 Meta-analysis 23 24 A similar protocol of the MA was previously published in the following studies 25 26 [13] [21]. ES will be calculated by Normalized Mean Difference and expressed Hedges’g 27 28 with the upper and lower 95% confidence limits. The statistical model of analysis is 29 30 31 Random-effects Model. Heterogeneity between studies will be assessed by using the Chi- 32 33 square statistic with n-1 degrees of freedom (DF). Subgroup analysis will cluster data by 34 http://openscience.bmj.com/ 35 species, strain, age, sex, type of molecular marker, time of administration of 36 37 38 antidepressants (acute, subacute, chronic), quality scores (use of randomization and blind 39 40 evaluation of the outcome). A meta-regression and, if necessary stratified, will be 41 42 performed to evaluate doses of the treatments. Funnel Plotting, Egger regression, trim and 43

44 on September 27, 2021 by guest. Protected copyright. 45 fill, and p Curve will be used to estimate the risk of publication bias. Calculations will be 46 47 performed using R Studio [22]. 48 49 Prospects: 50 51 The results of SR1, SR 2 and SR3 and the respective MA will be published 52 53 54 independently in a peer-reviewed scientific journal. Implementation of a Living 55 56 Systematic Review (LSR) in this protocol may be useful to keep the systematic reviews 57 58 continuously updated. That is, relevant studies will be incorporated into the previously 59 60

https://mc.manuscriptcentral.com/bmjos Page 17 of 21 BMJ Open Science

16 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 structured database of a standard systematic review as soon as they become available 4 5 [23][24] 6 . Tools in bibliographic platforms such as PubMed offer periodical reports on a 7 8 systematic review registered in the system. The LSR format is particularly important 9 10 when a research field generates new data frequently. 11 12 Confidential: For Review Only 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 http://openscience.bmj.com/ 36 37 38 39 40 41 42 43

44 on September 27, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 18 of 21

17 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 REFERENCES 4 5 6 1. Organization for Economic Cooperation and Development. Health at a 7 8 Glance 2015: OECD Indicators, Publishing, Paris, 2015. 9 10 2. Leonard, B. Clinical implications of mechanisms of action of antidepressants. 11 12 Confidential:Advances in Psychiatric Treatment For 2000;6(3): Review 178−86. Only 13 14 15 3. Liu B. From Serotonin to Neuroplasticity: Evolvement of Theories for Major 16 17 Depressive Disorder. Front Cell Neuroscience 2017;28(11):305. 18 19 4. Delgado P.L; D.S. Charney; L.H. Price, G.K. Aghajanian, H. Landis, G.R. 20 21 22 Heninger L.H. Serotonin function and the mechanism of antidepressant action: 23 24 Reversal of antidepressant-induced remission by rapid depletion of plasma 25 26 Archives General Psychiatry 1990;47:411-18. 27 28 5. DelgadoP.L; Delgado, H.M; Miller, R.M; Salomon, J; Licinio, A.J; Gelenberg, 29 30 31 D.S. Monoamines and the mechanism of antidepressant action: Effects of 32 33 catecholamine depletion on mood in patients treated with antidepressants 34 http://openscience.bmj.com/ 35 Psychopharmacology Bulletin 1993;29:389-96. 36 37 38 6. Malberg J.E; Eisch A.J; Nestler E.J; Duman R.S. Chronic antidepressant 39 40 treatment increases neurogenesis in adult rat hippocampus. Journal of 41 42 Neurosciences 2000; 20:9104 –10. 43

44 on September 27, 2021 by guest. Protected copyright. 45 7. Santarelli L; Saxe M; Gross C; Surget A; Battaglia F; Dulawa S; Weisstaub N; 46 47 Lee J; Duman R; Arancio O; Belzung C; Hen R. Requirement of hippocampal 48 49 neurogenesis for the behavioral effects of antidepressants. Science 50 51 2003;301(5634):805-9. 52 53 54 8. Miller B.R; Hen R. The current state of the neurogenic theory of depression and 55 56 anxiety. Curr Opin Neurobiol 2015:51–58. 57 58 59 60

https://mc.manuscriptcentral.com/bmjos Page 19 of 21 BMJ Open Science

18 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 9. Eliwa H; Belzung C; Surget A. Adult hippocampal neurogenesis: Is it the alpha 4 5 6 and omega of antidepressant action? Biochemical Pharmacology 2017; 141:86 – 7 8 99. 9 10 10. Surget A; Tanti A; Leonardo E.D; Laugeray A; Rainer Q et al. Antidepressants 11 12 Confidential:recruit new neurons to improve For stress responseReview regulation. Only Molecular Psychiatry 13 14 15 2011; 16:1177 – 88. 16 17 11. David D.J; Samuels B.A; Rainer Q; Wang J.W; Marsteller D; Mendez I et al. 18 19 Neurogenesis-dependent and -independent effects of fluoxetine in an animal 20 21 22 model of anxiety/depression. Neuron 2009; 62:479 – 93. 23 24 12. Nollet P.M; Tanti A; Girault V; Belzung C; Lema S. Neurogenesis-independent 25 26 antidepressant-like effects on behavior and stress axis response of a dual orexin 27 28 in a rodent model of depression. Neuropsychopharmacology 29 30 31 2012; 37:2210-21. 32 33 13. Lino de Oliveira C; Bolzan J.A; Surget A; Belzung C. Do antidepressants promote 34 http://openscience.bmj.com/ 35 neurogenesis in adult hippocampus? A systematic review and meta-analysis on 36 37 38 naive rodents. Pharmacology & Therapeutics 2020 (Available online) 39 40 10.1016/j.pharmthera.2020.107515. 41 42 14. Malberg J.E; Duman R.S; Cell proliferation in the adult hippocampus is decreased 43

44 on September 27, 2021 by guest. Protected copyright. 45 by inescapable stress: reversal by fluoxetine treatment. 46 47 Neuropsychopharmacology 2003;28(9):1562-71. 48 49 15. Ho N.F, Hooker J.M; Sahay A; Holt D.J; Roffman J.L. In vivo imaging of adult 50 51 human hippocampal neurogenesis: Progress, pitfalls and promise. Molecular 52 53 54 Psychiatry 2013; 18:404-16. 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 20 of 21

19 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 16. De Vries R.B.M; Hooijmans C.R; Langendam M.W et al. A protocol format for 4 5 6 the preparation, registration and publication of systematic reviews of animal 7 8 intervention studies. Evid Based Preclinical Medicine, 2015; 2: e00007–9. 9 10 17. Macleod MR, O’Collins T, Howells DW, et al. Pooling of animal experimental 11 12 Confidential:data reveals influence of study Fordesign and Review publication bias. Only Stroke 2004;35:1203- 13 14 15 1208. 16 17 18. Hooijmans, C.R., Rovers, M.M., de Vries, R.B. et al. SYRCLE’s risk of bias tool 18 19 for animal studies. BMC Med Res Methodol 2014, 14: 43. 20 21 22 https://doi.org/10.1186/1471-2288-14-43 23 24 19. Pico (Systematic Reviews - Research Guide), available online: link: 25 26 https://libguides.murdoch.edu.au/systematic/PICO. 27 28 29 20. Ramos-Hryb A.B; Bahor Z; McCann S et al. Protocol for a systematic review and 30 31 meta-analysis of data from preclinical studies employing forced swimming test: 32 33 an update. BMJ Open Science 2019;3: e 000035. doi:10.1136/ bmjos-2017- 34 http://openscience.bmj.com/ 35 000043. 36 37 38 21. AB Ramos-Hryb, C Harris, O Aighewi, C Lino-de-Oliveira. How would 39 40 publication bias distort the estimated effect size of prototypic antidepressants in 41 42 the forced swim test? Neuroscience & Biobehavioral Reviews 2018; 92: 192-194. 43

44 on September 27, 2021 by guest. Protected copyright. 45 22. R Development Core Team. An Introduction to R. R Foundation for Statistical 46 47 Computing, Vienna, Austria 2008. ISBN 3-900051-12-7 available in 48 49 https://www.r-project.org/about.html 50 51 52 23. Thomas et al. Living systematic reviews: 2. Combining human and machine 53 54 effort. Journal of Clinical Epidemiology 2017; 91:31-37. 55 56 24. Elliott et al. Living systematic review: 1. Introduction-the why, what, when, and 57 58 how. Journal of Clinical Epidemiology 2017; 91:23-30. 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from

Page 21 of 21 BMJ Open Science

1 2 3 4 5 6 7 8 9 10 11 Confidential: For Review Only 12 13 14 15 16 Timeline of a systematic review and meta-analysis. Note: Icons in the illustration are available in 17 flaticon.com 18 19 160x44mm (300 x 300 DPI) 20 21 22 23 24 25 26 27 28 29 30 http://openscience.bmj.com/ 31 32 33 34 35 36 37 38 39 40 on September 27, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmjos BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from

BMJ Open Science Page 22 of 21

1 2 3 4 5 6 7 8 9 10 11 Confidential: For Review Only 12 13 14 15 16 17 18 19 20 21 Types of primary studies included in every systematic review according to population, intervention, 22 comparison and outcome.Note: Icons in the illustration are available in flaticon.com 23 24 170x74mm (220 x 220 DPI) 25 26 27 28 29 30 http://openscience.bmj.com/ 31 32 33 34 35 36 37 38 39 40 on September 27, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmjos BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from

BMJ Open Science

Confidential: For Review Only A protocol for systematic review and meta-analysis of the evidence linking hippocampal neurogenesis to the effects of antidepressants on mood and behaviour

Journal: BMJ Open Science

Manuscript ID bmjos-2020-100077.R1

Article Type: Protocol

Date Submitted by the 18-Sep-2020 Author:

Complete List of Authors: Bolzan, Juliana; Universidade Federal de Santa Catarina, Departamento de Ciências Fisiológica; Universidade Federal de Santa Catarina, Pós- Graduação em Farmacologia Lino de Oliveira, Cilene; Universidade Federal de Santa Catarina, Departamento de Ciências Fisiológicas; Universidade Federal de Santa Catarina, Pós-Graduação em Farmacologia

Keywords: http://openscience.bmj.com/

on September 27, 2021 by guest. Protected copyright.

https://mc.manuscriptcentral.com/bmjos Page 1 of 22 BMJ Open Science BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 4 5 6 7 8 9 I, the Submitting Author has the right to grant and does grant on behalf of all authors of the Work (as defined 10 in the below author licence), an exclusive licence and/or a non-exclusive licence for contributions from authors 11 who are: i) UK Crown employees; ii) where BMJ has agreed a CC-BY licence shall apply, and/or iii) in accordance 12 with the termsConfidential: applicable for US Federal Government For officers Review or employees actingOnly as part of their official 13 duties; on a worldwide, perpetual, irrevocable, royalty-free basis to BMJ Publishing Group Ltd (“BMJ”) its 14 licensees and where the relevant Journal is co-owned by BMJ to the co-owners of the Journal, to publish the 15 Work in this journal and any other BMJ products and to exploit all rights, as set out in our licence. 16 17 The Submitting Author accepts and understands that any supply made under these terms is made by BMJ to 18 the Submitting Author unless you are acting as an employee on behalf of your employer or a postgraduate 19 student of an affiliated institution which is paying any applicable article publishing charge (“APC”) for Open 20 Access articles. Where the Submitting Author wishes to make the Work available on an Open Access basis (and 21 intends to pay the relevant APC), the terms of reuse of such Open Access shall be governed by a Creative 22 Commons licence – details of these licences and which Creative Commons licence will apply to this Work are set 23 out in our licence referred to above. 24 25 Other than as permitted in any relevant BMJ Author’s Self Archiving Policies, I confirm this Work has not been 26 accepted for publication elsewhere, is not being considered for publication elsewhere and does not duplicate 27 material already published. I confirm all authors consent to publication of this Work and authorise the granting 28 of this licence. 29 30 31 32 33 34 35 http://openscience.bmj.com/ 36 37 38 39 40 41 42 43

44 on September 27, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 2 of 22

1 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 Title: A protocol for systematic review and meta-analysis of the evidence linking 4 5 6 hippocampal neurogenesis to the effects of antidepressants on mood and behaviour 7 8 Authors: Juliana A Bolzan 1,2, Cilene Lino de Oliveira1,2. 9 10 11 Affiliations: 12 Confidential: For Review Only 13 14 1- Department of Physiological Sciences, Center of Biological Sciences, Federal 15 16 University of Santa Catarina – UFSC, Campus Trindade, 88040-900, Florianópolis - SC, 17 18 Brazil. 19 20 21 2- Post graduation in Pharmacology, Center of Biological Sciences, Federal University 22 23 24 of Santa Catarina – UFSC, Campus Trindade, 88040-900, Florianópolis - SC, Brazil. 25 26 Address to correspondence: 27 28 29 Cilene Lino de Oliveira 30 31 32 LABORATORY OF BEHAVIORAL NEUROBIOLOGY (LABNEC-CFS-UFSC)- 33 34

Departmento de Ciências Fisiológicas, Centro de Ciências Biológicas, Universidade http://openscience.bmj.com/ 35 36 37 Federal de Santa Catarina. Campus Universitário, Trindade, CEP: 88040-900, 38 39 Florianópolis, SC – Brazil, Phone: +55 48 3721-7085, E-mail: [email protected], 40 41 Orcid: orcid.org/0000-0002-0627-530X 42 43

44 Running title: Hippocampal neurogenesis and antidepressants. on September 27, 2021 by guest. Protected copyright. 45 46 Word counting (except the title page, abstract, references, figures, tables or 47 48 49 acknowledgements): 2.900 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos Page 3 of 22 BMJ Open Science

2 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 Authors’ contributions: 4 5 6 7 CLO: selected research theme; performed preliminary studies; wrote the preliminary 8 9 protocol; wrote the manuscript; approved final version of the manuscript. 10 11 JAB: performed preliminary studies; wrote the preliminary protocol; wrote the 12 Confidential: For Review Only 13 14 manuscript; approved the final version of the manuscript. 15 16 Funding: This study was financed in part by the Coordenação de Aperfeiçoamento de 17 18 Pessoal de Nível Superior – Brasil (CAPES) – Finance Code 001. Juliana A Bolzan is a 19 20 recipiente of a fellowship from Conselho Nacional de Desenvolvimento Científico e 21 22 23 Tecnológico, Brazil (131247/2019-0). 24 25 Competing interests: Senior author (Cilene Lino de Oliveira) is an Associate Editor of 26 27 BMJ Open Science. Other than that, no conflict of interest is declared. 28 29 30 Open data: open data are available at (https://osf.io/5z8us/) 31 32 33 34 35 http://openscience.bmj.com/ 36 37 38 39 40 41 42 43

44 on September 27, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 4 of 22

3 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 ABSTRACT 4 5 6 Objective: Studies in rodents associated the deficits of adult hippocampal neurogenesis 7 8 with behavioural anomalies which may be reversed by antidepressant treatments. A 9 10 previous systematic review (SR) and meta-analysis (MA) indicated a hierarchy within the 11 12 pro neurogenicConfidential: effects of different antidepressants For Review in naive rodents. Only The present review 13 14 15 aims to evaluate a more comprehensive sample of studies investigating the links between 16 17 the effects of different antidepressants and adult hippocampal neurogenesis. Search 18 19 strategy, Screening annotation, Data management: Protocols were planned following 20 21 22 PRISMA-P guidelines. Searches in Embase, Medline, Scopus and Web of Science 23 24 followed by screening with inclusion/exclusion criteria will provide relevant publications. 25 26 First SR will summarize the effects of antidepressants on adult hippocampal neurogenesis 27 28 on different laboratory rodents. Second SR will summarize the correlations between 29 30 31 neurogenic and behavioural effects of antidepressants while the third will focus on cause- 32 33 effect relationships between them. If feasible, data will be analyzed by pairwise or 34 http://openscience.bmj.com/ 35 network random-effect or multivariate MA to determine the direction, magnitude, 36 37 38 significance and heterogeneity (I-squared statistics) of the estimated effect sizes on global 39 40 or subgroup levels. Funnel plotting, Egger regression, 'trim and fill' will be used to 41 42 estimate the risk of publication bias. Quality assessment of the included publications will 43

44 on September 27, 2021 by guest. Protected copyright. 45 be performed by applying adapted CAMARADES, Syrcles' RoB or CINeMA tools. 46 47 Reporting: Find a preliminary version of this protocol at the Open Science Framework 48 49 (https://osf.io/gmsvd/). Data extraction has already started. Results shall be published in 50 51 a peer-reviewed journal. Due to the continuous production in the field, the 52 53 54 implementation of a "Living Systematic Review" is intended. 55 56 Keywords: hippocampus, neurogenesis, antidepressants, effect size, heterogeneity, 57 58 animal models. 59 60

https://mc.manuscriptcentral.com/bmjos Page 5 of 22 BMJ Open Science

4 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 STRENGTHS OF THIS STUDY 4 5  6 A preliminary version of this protocol is registered at Open Science Framework. 7 8  Search and screening cover an extensive range of publications. 9 10  Two independent reviewers will perform data extraction. 11 12  Confidential:Quality of studies will be assessed For with Reviewthree different tools. Only 13 14 15  Meta-analyses are planned to include indirect comparisons and non-dependent ES. 16 17 LIMITATIONS OF THIS STUDY 18 19  The protocol is not registered at PROSPERO because data extraction already 20 21 22 began. 23 24  Reviewers will be not blind to the bibliographic information of primary studies. 25 26  A single reviewer performed search and screening, a second one double-checked. 27 28  Correlation between subgroups of studies may not be discarded. 29 30 31 32 33 34 35 http://openscience.bmj.com/ 36 37 38 39 40 41 42 43

44 on September 27, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 6 of 22

5 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 INTRODUCTION 4 5 6 7 The consumption of antidepressant medications used in psychiatric treatments 8 9 doubled according to the Organization of Economic Cooperation and Development 10 11 (OECD) between the years 2000 and 2015 [1]. Despite a large number of patients under 12 Confidential: For Review Only 13 treatment, the mechanisms underlying therapeutic effects of antidepressants remain 14 15 16 elusive. Behavioural studies indicate that the therapeutic effects of antidepressants 17 18 depend on the reversal of neurochemical deficiencies [2]. For example, earlier studies 19 20 revealed that most antidepressants inhibit the reuptake of monoamines, promoting high 21 22 [3] 23 extracellular concentrations in the central nervous system . Moreover, antagonism of 24 25 monoaminergic transmission provoked sudden decay of the effectiveness of 26 27 antidepressants in patients [4, 5]. Although durable, the relationships between monoamines, 28 29 depression and antidepressants may not explain all aspects of antidepressant therapy. For 30 31 32 example, the delay between the onset of treatment and remission of symptoms or the 33 34 existence of depressions resistant to prototype antidepressants suggests there are more for 35 http://openscience.bmj.com/ 36 depression than monoamines. 37 38 39 Animal models allow for discoveries generating hypothesis and theories on 40 41 depression and antidepressant efficacy beyond monoamines. Neurogenic theory of 42 43 depression gained more attention in the 2000s when a series of papers reported that 44 on September 27, 2021 by guest. Protected copyright. 45 46 chronic antidepressant treatment increased the number of newborn neurons in the 47 48 hippocampus of adult rodents [6,7,8] suggesting a pro neurogenic effect of the treatment. 49 50 Stressful stimuli provoked hippocampal cell loss and behavioural sequelae prevented by 51 52 chronic antidepressant treatment in rats [9] indicating a positive correlation between pro 53 54 55 neurogenic and behavioural effects of the compounds. Furthermore, ablation of 56 57 hippocampal neurogenesis reverted the effects of antidepressants on the behaviour of 58 59 adult mice [7,8,10] suggesting a cause-effect relationship between pro neurogenic actions 60

https://mc.manuscriptcentral.com/bmjos Page 7 of 22 BMJ Open Science

6 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 of antidepressants and their effects on behaviour. Altogether, these data engendered an 4 5 6 explanation considering hippocampal neurogenesis as the underlying mechanism for the 7 8 efficacy of antidepressants on treating mood and behavioural disorders. 9 10 Since the neurogenesis is a time-consuming process, the neurogenic theory offers 11 12 a plausibleConfidential: explanation to delay between For the onsetReview of antidepressant Only treatment and the 13 14 15 therapeutic effect. Neurogenic theory inspired a long series of studies aiming to 16 17 understand details of the process which could provide a variety of new targets for the 18 19 development of antidepressants. Over time the findings diverging of the earliest 20 21 22 publications emerged indicating that some effects of antidepressants could be related to 23 24 hippocampal neurogenesis while others would be unrelated to it [11,12]. For example, the 25 26 effects of chronic antidepressants on behavioural tests such as novelty suppressed feeding 27 28 seem to correlate positively and depend on hippocampal neurogenesis [7]. In contrast, the 29 30 31 effects of antidepressants on the forced swimming test seem unrelated to it [11]. A pioneer 32 33 systematic review and meta-analysis of studies investigating effects of antidepressants in 34 http://openscience.bmj.com/ 35 the hippocampus of adult laboratory rodents suggested that some compounds increased 36 37 [13] 38 the neurogenesis while others did not . Depending on the type of the compound 39 40 (fluoxetine, imipramine, venlafaxine, desvenlafaxine), laboratory species (rat or mice) 41 42 and the type of molecular marker (BrdU, doublecortin, Ki-67), the magnitude of the effect 43 44 [13] on September 27, 2021 by guest. Protected copyright. 45 size (ES) may vary from negligible to very large . The broad range of ES values 46 47 indicates that actions of antidepressants may be more or less related or dependent on 48 49 hippocampal neurogenesis. 50 51 To further explore the concept of the hierarchy of effect sizes within different 52 53 54 antidepressants, present work proposes to analyze the data of hippocampal neurogenesis 55 56 with network meta-analysis. Network meta-analysis may enable indirect comparisons 57 58 between effect sizes [14]. Additionally, the present protocol also plans to include studies 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 8 of 22

7 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 of all times to perform a comprehensive systematic review and meta-analysis, which may 4 5 [13] 6 help to overcome some of the limitations of the previous one . Although enough 7 8 powered to evaluate effect sizes of different antidepressants on different species of 9 10 laboratory rodents, boundaries of dates (five years between 2013-2018) may have reduced 11 12 the amountConfidential: of data available to assess For other sourcesReview of heterogeneity Only [13]. Indeed, the 13 14 15 information was insufficient to investigate the influence of relevant characteristics of the 16 17 population such as sex, age and co-morbidities (stressed, anxious, depressed phenotypes) 18 19 [13]. Scarcity of relevant data also precluded the stratified analysis of the effect sizes of 20 21 [13] 22 the antidepressants by doses, via of administration or efficacy in behavioural testing . 23 24 Considering there is a variety of primary studies investigating the theme, it seems 25 26 not feasible to obtain all relevant information with a single systematic review. In a pilot 27 28 analysis, at least three categories of studies investigating the links between hippocampal 29 30 31 neurogenesis and antidepressants in laboratory rodents were identified in the literature. 32 33 For convenience, these categories were here named as 1- interventional, 2- correlational, 34 http://openscience.bmj.com/ 35 3-functional. Interventional category comprised those studies on the effects of the 36 37 38 intervention (antidepressant) on the different markers of the hippocampal neurogenesis 39 40 (neurogenic outcome), for example [6,7,8]. The correlational category included the studies 41 42 on the effects of the antidepressant (intervention) on a behavioural testing (outcome) and 43 44 [15,16] on September 27, 2021 by guest. Protected copyright. 45 different phases of the hippocampal neurogenesis (outcome), for instance . 46 47 Functional studies were those making ablation of hippocampal neurogenesis 48 49 (intervention), co-treated or not with antidepressants, before behavioural testing 50 51 (outcome), for example [7]. Therefore, three different systematic reviews (SR) were 52 53 54 planned to summarize interventional (SR 1), correlational (SR 2) and functional (SR 3) 55 56 studies. In summary, SR 1 will further extend and update previous meta-analysis [13] while 57 58 SR 2 and SR 3 will explore newer aspects of the subject. 59 60

https://mc.manuscriptcentral.com/bmjos Page 9 of 22 BMJ Open Science

8 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 The “neurogenic outcome”, appraised in SR 1 and 2, often means the combination 4 5 6 of different outcomes (measures of stemness, proliferation, survival, maturation of 7 8 newborn neurons). Besides, every behavioural testing, evaluated in SR 2 and 3, should 9 10 provide a different behavioural outcome (e.g. immobility time in the forced swimming 11 12 test, latencyConfidential: to touch the food in the noveltyFor suppressed Review feeding, Onlyso on). Since random- 13 14 15 effects model assumes independence of the effect sizes, but often information from the 16 17 primary studies did not allow to discard correlations between them, current protocol plan 18 19 to apply multivariate meta-analysis to take non-independence of data into account [17]. 20 21 22 23 METHODS AND ANALYSIS 24 25 26 Overview of the systematic reviews and meta-analyses. 27 28 29 The protocol and the manuscript follow PRISMA-P statement and Syrcle’s 30 31 formulary. However, approval by the ethics committee and the Helsinki Declaration does 32 33 34 not apply. The plan of every systematic review (SR 1, SR 2, SR 3) included the following 35 http://openscience.bmj.com/ 36 steps: 1-research question; 2-choice of search strategy; 3-choice of bibliographic 37 38 databases; 4- choice of inclusion and exclusion criteria; 5- planning of data extraction and 39 40 41 6- planning of the meta-analysis (Figure 1). SR 1, 2 and 3 aim to answer the following 42 43 questions prepared using the PICO tool [18], respectively: 1-do different antidepressants

44 on September 27, 2021 by guest. Protected copyright. 45 increase hippocampal neurogenesis at different degrees in laboratory rodents? 2-are there 46 47 48 correlations between behavioural and neurogenic effects of antidepressants in laboratory 49 50 rodents? 3- is there a causal link between behavioural and neurogenic effects of 51 52 antidepressants in laboratory rodents? 53 54 In the three reviews, the population comprises laboratory rodents independent of 55 56 57 species, strain, age, sex or co-morbidities. Interventions and outcomes vary across SR, as 58 59 follows. The SR 1 includes all studies investigating the effects of antidepressant treatment 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 10 of 22

9 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 (intervention) on adult hippocampal neurogenesis (outcome), independent of the type of 4 5 6 experimental design or association with behavioural testing. SR 2 will summarize the 7 8 correlation between outcomes (hippocampal neurogenesis and behaviour). In those 9 10 studies, the effects of antidepressant treatment (intervention) on adult hippocampal 11 12 neurogenesisConfidential: (outcome) and behaviour For testing Review (outcome) are investigated Only in the same 13 14 15 cohort of laboratory animals. Values of correlation will be extracted from primary studies, 16 17 and when absent, calculated from the extracted data. SR 3 will condense those studies 18 19 investigating the effects of ablating hippocampal neurogenesis (intervention) in 20 21 22 laboratory rodent, with or without co-treatment with an antidepressant, evaluated in the 23 24 selected behaviour testing (outcome). 25 26 Neurogenesis will be assessed through the measures of different molecular 27 28 markers of neurogenesis in the adult hippocampus (BrdU, Ki-67, DCX), arbitrarily 29 30 31 chosen according to previously published SRMA [13]. The different stages of neurogenesis 32 33 (Figure 2) will be estimated by quantification of the following markers on the dentate 34 http://openscience.bmj.com/ 35 gyrus of the hippocampus: 1- Nestin (expressed in stem cells), 2- BrdU (incorporated by 36 37 38 proliferating cell) or Ki-67 (expressed in proliferating cells, or other markers such as 39 40 Tbr2, Mash 1, PCNA), 3- DCX (expressed in immature neurons), 4- BrdU and NeuN 41 42 (expressed in mature neurons or other markers of mature neurons such as Map-2, Fox-3, 43

44 on September 27, 2021 by guest. Protected copyright. 45 Prox-1) co-localized (indicating mature neurons incorporated BrdU and matured in the 46 47 adulthood). Measurements of the markers may be using stereological or non-stereological 48 49 methods. The behavioural outcomes depend on the type of behavioural test reported (e.g., 50 51 time of immobility in forced swimming or tail-suspension test, latency to eat in the 52 53 54 novelty suppressed feeding, so on). Data extract from each SR will be analyzed 55 56 separately. Analytical choices were made taken characteristics of the outcomes into 57 58 account. 59 60

https://mc.manuscriptcentral.com/bmjos Page 11 of 22 BMJ Open Science

10 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 4 5 6 7 Search strategy 8 9 10 The search strategies (see Supplementary Table 1, Supplementary Table 2 and 11 12 SupplementaryConfidential: Table 3 at the Open Science For Framework, Review access link: Only https://osf.io/tajhf/) 13 14 were conceived after pilot studies using combinations of keywords from protocols 15 16 [19,20;21] 17 previously published . Databases for the searches (“Pubmed”, “Scopus”, 18 19 “Embase”, “Web of Science”) were arbitrarily chosen considering their extensive 20 21 collection of international publishers in the field of basic biomedicine and our institutional 22 23 24 access. The keywords for SR 1 (Supplementary Table 1) will be a combination of MeSH 25 26 terms related to the primary outcome (hippocampal neurogenesis, #1-5 ‘neurogenesis’ 27 28 and ‘hippocampus’ and (or ‘BrdU’ or ‘DCX’ or ‘Ki-67’) and intervention 29 30 (antidepressants, #6). The keywords for the SR 2 will be terms related to behavioural 31 32 33 outcomes (#7 Supplementary Table 2) combined with the terms of SR 1 (Supplementary 34 35 Table 1). The search terms exclusive of the SR 2 were chosen by examining the http://openscience.bmj.com/ 36 37 publications obtained in SR 1. The keywords for the SR 3 will be terms related to the 38 39 40 ablation of neurogenesis (#8, Supplementary Table 3) combined with the terms related to 41 42 the behavioural outcomes (#7, Supplementary Table 2) and neurogenic outcomes (#1-5, 43 44 Supplementary Table 1). Terms exclusive of SR 3 were chosen by examining the on September 27, 2021 by guest. Protected copyright. 45 46 [7] 47 references similar to Santarelli et al., 2003 . Publications returned of every search will 48 49 be exported to a reference manager file of EndNote X7. Searches were performed by one 50 51 reviewer (JB) and will be double checked for a second reviewer. 52 53 54 Screening processes: 55 56 57 58 Screening of publications returned from the searches will occur in two rounds, the 59 60 first one analyzing title and abstract and second-round evaluating full text. The exclusion

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 12 of 22

11 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 criteria are reviews, systematic reviews, meta-analyses, duplicates (all SR) and co- 4 5 6 treatment (SR 1 and 2). The first round will exclude reviews, systematic reviews, meta- 7 8 analyses, duplicates. The second round of screening will be required to identify co- 9 10 treatments, exclusion criterion in SR 1 and 2, and apply inclusion criteria for all SRs. 11 12 Confidential:All controlled studies will be For included Review in SRs, regardless Only of randomization, 13 14 15 blinding or study design (Figure 3). There will have no restriction on language or dates. 16 17 Studies on adult rats or mice, regardless of sex, strain or stress background; studies with 18 19 any route, dose and treatment schedule for drug administration; studies measuring any 20 21 22 molecular markers in the dentate gyrus of the hippocampus (see Figure 3) will be eligible. 23 24 In these studies, interventions (antidepressants in SR 1 and 2; ablation of neurogenesis 25 26 with or without co-treatment with an antidepressant, SR 3) will be compared to controls, 27 28 i.e., vehicle (SR 1 and 2) or sham ablation or the absence of the intervention (SR 3). 29 30 31 Although studies with co-treatments will be excluded in SR 1 and 2, if the same study 32 33 includes groups of animals treated with a single antidepressant and the respective control 34 http://openscience.bmj.com/ 35 group, these groups will be included. One reviewer performed the screening process; a 36 37 38 second reviewer will double-check the screening. Systematic reviewers will not be 39 40 blinded to authors, date or journal of primary articles. 41 42 43 Quality of study 44 on September 27, 2021 by guest. Protected copyright. 45 46 47 Quality of every publication will be estimated by the use of an adapted version of 48 49 CAMARADES checklist [22] (Table 1) and RoB Syrcle tools [23] which assess, 50 51 respectively, general aspects of experimental design and more specific features of animal 52 53 research. CINeMA tool will be used in the case of the implementation of the network 54 55 56 meta-analysis [24]. 57 58 The quality of every study evaluated with adapted CAMARADES checklist will 59 60 vary from score zero (minimum) to ten (maximum). Every study may receive a score

https://mc.manuscriptcentral.com/bmjos Page 13 of 22 BMJ Open Science

12 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 when reporting information required in the checklist anywhere in the text (title, abstract 4 5 6 or body of the text). This version of the CAMARADES checklist was adapted using items 7 8 6 and 9 of ARRIVE 2.0 guidelines [25]. A score at item 9 may be obtained by reporting 9 10 (ARRIVE item 9): "a. What was done, how it was done, and what was used? b. When 11 12 and howConfidential: often. c. Where (including For detail ofReview any acclimatization Only periods). d. Why 13 14 15 (provide a rationale for procedures)". A score at item 10 may be obtained by reporting 16 17 (ARRIVE item 6): “Clearly define all outcome measures assessed (e.g., cell death, 18 19 molecular markers, or behavioural changes)”. The quality of every study evaluated using 20 21 22 Syrcle's RoB tool will be performed by answering the ten signalling questions with "yes", 23 24 "unclear" or "no", which will be considered a "low", "unclear" and "high" risk of bias, 25 26 respectively [23]. The quality using CINeMA tool [24] considers the following domains: (i) 27 28 within-study bias, (ii) reporting bias, (iii) indirectness, (iv) imprecision, (v) heterogeneity, 29 30 31 and (vi) incoherence. To each domain, threes levels o judgement may be assigned: 1-no 32 33 concerns, 2-some concerns, or 3-major concerns. Judgments may be summarised to 34 http://openscience.bmj.com/ 35 obtain four levels of confidence: 1-very low, 2-low, 3-moderate, or 4-high. Two 36 37 38 independent reviewers will perform the quality assessment. A third reviewer will solve 39 40 discrepancies. 41 42 Table 1 – Adapted CAMARADES' study quality checklist 43

44 1-peer-reviewed publication. on September 27, 2021 by guest. Protected copyright. 45 2- studies following ARRIVE (or other) guidelines. 46 3- declaration of compliance with animal testing regulations and legislation. 47 48 4- declaration of interest. 49 5- report of the breeding, husbandry conditions and actions to improve animal welfare 50 of the experimental animals (for example, environmental enrichment). 51 52 6-report of the species, lineage or other identifying characteristics of the experimental 53 animals (e.g. types of transgenes or knockouts) or phenotypes of interest (Stressed? 54 Depressed? Anxious? Other?). 55 56 7- report of the age, weight or stage of the experimental animals. 57 8- report of the sex of the experimental animals. 58 59 9- report of the methods of behavioural testing and acquisition of the behavioural 60 outcomes.

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 14 of 22

13 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 10- report of the methods to the detection of molecular markers, including the brand of 4 5 antibodies and kits, and the acquisition of the molecular outcomes. 6 7 8 9 10 Outcome extraction 11 12 Confidential: For Review Only 13 14 For a qualitative and quantitative summary of the included studies, bibliographic 15 16 information and data on the experimental design, population, intervention, comparison, 17 18 and the qualitative and quantitative aspects of the outcomes will be extracted from every 19 20 publication (Table 2). Qualitative data will be obtained from the text or tables. 21 22 23 Quantitative data will be obtained from the text or tables of publications or using a digital 24 25 ruler to extract them from the charts. When information is entirely unavailable in the 26 27 publication, it may request directly by contact with the authors by e-mail of the 28 29 30 corresponding author. In the absence of a response, within 30 days, the study will be 31 32 excluded from the analysis. Two independent reviewers will extract data. A third reviewer 33 34

will solve discrepancies. Every independent reviewer will input qualitative and http://openscience.bmj.com/ 35 36 37 quantitative data extracted from publications in an Excel sheet structured with a study per 38 39 row and a variable per column. Agreement between reviewers will be assessed by using 40 41 “comparing columns” tool of the Excel (qualitative data) or by calculating Cohen Kappa 42 43 index (quantitative data). Discrepancies above 20 %, i.e., a minimum 80% of concordance 44 on September 27, 2021 by guest. Protected copyright. 45 [26] 46 considered substantial , will be conciliated by the third reviewer. 47 48 Table 2 − Data extracted from each publication 49 50 Bibliographic information Authorship, year, journal 51 Experimental design See items in the adapted CAMARADES 52 53 checklist (Table 1) and RoB Syrcle tool 54 [23] 55 Population Species, strain, sex, age, the timing of 56 57 euthanasia, phenotype of experimental 58 rodents 59 60

https://mc.manuscriptcentral.com/bmjos Page 15 of 22 BMJ Open Science

14 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 Intervention SR 1-3 (antidepressants): Type, 4 mechanism of action, dose, via, number of 5 6 injections per day. 7 SR 3 (ablation of neurogenesis): the 8 method employed to the ablation 9 Comparison SR 1,2: Type of vehicle. 10 SR 3: the method employed to the sham 11 ablation. 12 Confidential: For Review Only 13 Outcome (qualitative) SR 1,2: Type of marker, antibodies used, 14 the protocol for detection, protocols for 15 BrdU injection (survival or proliferation), 16 protocols of measuring (stereological or 17 non-stereological). SR 2,3: type, metrics 18 19 and method of acquisition (manual or 20 automatic) of the behavioural outcome. 21 Outcome (quantitative) All SRs (per outcome, for experimental 22 and control groups): mean, standard 23 24 deviation or standard error, sample size 25 (when available, effect sizes with 95% 26 confidence interval), p values or the 27 28 values of F or t, and correlation indexes 29 (SR 2). 30 31 Meta-analysis 32 33 34 35 Quantitative data extracted from publications (Table 2) will be used in the meta- http://openscience.bmj.com/ 36 37 analysis. A similar protocol of the MA was previously published in the following studies 38 39 [13] [27]. Data of each SR will be analyzed independently. For every study included in each 40 41 SR, an ES will be calculated by using a Normalized Mean Difference and expressed 42 43

44 Hedges’g with the upper and lower 95% confidence limits. For each SR, a global estimate on September 27, 2021 by guest. Protected copyright. 45 46 of ES will include all laboratory animals, all antidepressants and all markers for 47 48 neurogenesis. In the case of independent outcomes, random-effect models will be used 49 50 51 for the meta-analyses, and heterogeneity will be assessed by using the I-squared statistic. 52 53 Since the random-effects model assumes independence of the ES, when the primary 54 55 studies do not allow to discard correlations between outcomes, multivariate meta-analysis 56 57 58 will be performed. Multivariate meta-analysis may be used to estimate ES and 59 60 heterogeneity taking non-independence of data into account [17].

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 16 of 22

15 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 Stratified meta-analyses into the subgroups 1-population, 2-intervention and 3- 4 5 6 outcome are planned (Table 3). Meta-regressions and, if necessary stratified, will be 7 8 performed to evaluate doses of the antidepressants. For comparisons between ES of 9 10 different antidepressants, network meta-analyses are planned for data of each SR [14]. 11 12 FunnelConfidential: Plotting, Egger regression, trim For and fill, Review and p-Curve will beOnly used to estimate the 13 14 [28] 15 risk of publication bias. Calculations will be performed using R Studio . 16 17 18 Table 3 – Subgroup meta-analysis 19 Population Species, strain, sex, age, phenotype, stress 20 21 condition and behavioural testing. 22 Intervention Type, mechanism of action, dose, via, 23 number of injections per day. 24 25 Outcome SR 1,2: Type of marker, antibodies used, 26 the protocol for detection, protocols for 27 BrdU injection (survival or proliferation), 28 protocols of measuring (stereological or 29 non-stereological). SR 2,3: type, metrics 30 and method of acquisition (manual or 31 automatic) of the behavioural outcome. 32 33 34 Prospects: 35 http://openscience.bmj.com/ 36 37 The results of SR 1, SR 2 and SR 3 and the respective MA will be published 38 39 independently in a peer-reviewed scientific journal. Implementation of a Living 40 41 Systematic Review (LSR) in this protocol may be useful to keep the systematic reviews 42 43

44 continuously updated. That is, relevant studies will be incorporated into the previously on September 27, 2021 by guest. Protected copyright. 45 46 structured database of a formal systematic review as soon as they become available [29][30]. 47 48 Tools in bibliographic platforms such as PubMed offer periodical reports on a systematic 49 50 51 review registered in the system. The LSR format is particularly important when a research 52 53 field generates new data frequently. We expect that researchers interested in 54 55 antidepressant research, especially those worried with mechanisms of action, may benefit 56 57 58 from having updated information on this subject to plan new studies or make new 59 60 synthesis to improve theoretical models in the field.

https://mc.manuscriptcentral.com/bmjos Page 17 of 22 BMJ Open Science

16 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 REFERENCES 4 5 6 7 1. Organization for Economic Cooperation and Development. Health at a 8 Glance 2015: OECD Indicators, Publishing, Paris, 2015. 9 10 11 12 2.Confidential: Leonard, B. Clinical implications For of mechanismsReview of action Only of antidepressants. 13 Advances in Psychiatric Treatment 2000;6(3): 178−86. 14 15 16 17 3. Liu B. From Serotonin to Neuroplasticity: Evolvement of Theories for Major 18 19 Depressive Disorder. Front Cell Neuroscience 2017;28(11):305. 20 21 22 4. Delgado P.L; D.S. Charney; LH Price, G.K. Aghajanian, H. Landis, G.R. 23 24 Heninger L.H. Serotonin function and the mechanism of antidepressant action: 25 Reversal of antidepressant-induced remission by rapid depletion of plasma 26 27 tryptophan Archives General Psychiatry 1990;47:411-18. 28 29 30 31 5. Delgado P.L; Delgado, H.M; Miller, R.M; Salomon, J; Licinio, A.J; Gelenberg, 32 D.S. Monoamines and the mechanism of antidepressant action: Effects of 33 34 catecholamine depletion on mood in patients treated with antidepressants 35 http://openscience.bmj.com/ 36 Psychopharmacology Bulletin 1993;29:389-96. 37 38 39 6. Malberg J.E; Eisch A.J; Nestler E.J; Duman R.S. Chronic antidepressant treatment 40 41 increases neurogenesis in adult rat hippocampus. Journal of Neurosciences 2000; 42 43 20:9104 –10.

44 on September 27, 2021 by guest. Protected copyright. 45 46 7. Santarelli L; Saxe M; Gross C; Surget A; Battaglia F; Dulawa S; Weisstaub N; 47 48 Lee J; Duman R; Arancio O; Belzung C; Hen R. Requirement of hippocampal 49 50 neurogenesis for the behavioral effects of antidepressants. Science 51 2003;301(5634):805-9. 52 53 54 55 8. Miller B.R; Hen R. The current state of the neurogenic theory of depression and 56 57 anxiety. Curr Opin Neurobiol 2015:51–58. 58 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 18 of 22

17 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 9. Eliwa H; Belzung C; Surget A. Adult hippocampal neurogenesis: Is it the alpha 4 5 and omega of antidepressant action? Biochemical Pharmacology 2017; 141:86 – 6 7 99. 8 9 10 10. Surget A; Tanti A; Leonardo E.D; Laugeray A; Rainer Q et al. Antidepressants 11 12 Confidential:recruit new neurons to improve For stress responseReview regulation. Only Molecular Psychiatry 13 14 2011; 16:1177 – 88. 15 16 17 11. David D.J; Samuels B.A; Rainer Q; Wang J.W; Marsteller D; Mendez I et al. 18 19 Neurogenesis-dependent and -independent effects of fluoxetine in an animal 20 model of anxiety/depression. Neuron 2009; 62:479 – 93. 21 22 23 24 12. Nollet P.M; Tanti A; Girault V; Belzung C; Lema S. Neurogenesis-independent 25 26 antidepressant-like effects on behavior and stress axis response of a dual orexin 27 receptor antagonist in a rodent model of depression. Neuropsychopharmacology 28 29 2012; 37:2210-21. 30 31 32 33 13. Lino de Oliveira C; Bolzan J.A; Surget A; Belzung C. Do antidepressants promote 34

neurogenesis in adult hippocampus? A systematic review and meta-analysis on http://openscience.bmj.com/ 35 36 naive rodents. Pharmacology & Therapeutics 2020 (Available online) 37 38 10.1016/j.pharmthera.2020.107515. 39 40 41 14. Furukawa T.A; Salanti G; Atkinson L.Z, et al. Comparative efficacy and 42 43 acceptability of first-generation and second-generation antidepressants in the

44 on September 27, 2021 by guest. Protected copyright. 45 acute treatment of major depression: protocol for a network meta-analysis BMJ 46 Open 2016;6: e010919. 47 48 49 50 15. Malberg J.E; Duman R.S; Cell proliferation in the adult hippocampus is decreased 51 by inescapable stress: reversal by fluoxetine treatment. 52 53 Neuropsychopharmacology 2003;28(9):1562-71. 54 55 56 57 16. Ho NF, Hooker JM; Sahay A; Holt D.J; Roffman J.L. In vivo imaging of adult 58 human hippocampal neurogenesis: Progress, pitfalls and promise. Molecular 59 60 Psychiatry 2013; 18:404-16.

https://mc.manuscriptcentral.com/bmjos Page 19 of 22 BMJ Open Science

18 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 4 5 17. Van den Noortgate W; López-López J.A; Marín-Martínez F, et al. Meta-analysis 6 7 of multiple outcomes: a multilevel approach. Behav Res 2015; 47:1274-1294. 8 9 10 18. Pico (Systematic Reviews - Research Guide), available online: link: 11 12 Confidential:https://libguides.murdoch.edu.au/systematic/PICO. For Review Only 13 14 15 19. Ramos-Hryb A.B; Bahor Z; McCann S et al. Protocol for a systematic review and 16 17 meta-analysis of data from preclinical studies employing forced swimming test: 18 19 an update. BMJ Open Science 2019;3: e 000035. doi:10.1136/ bmjos-2017- 20 000043. 21 22 23 24 20. McCann S.K, Irvine C, Mead G.E, et al. Efficacy of antidepressants in animal 25 26 models of ischemic stroke: a systematic review and meta-analysis. Stroke 2014; 27 45:3055–63. 28 29 30 31 21. Lazic S.E; Fuss J; Gass P. Quantifying the Behavioural Relevance of 32 33 Hippocampal Neurogenesis. PLoS ONE 2014; 9:11. 34

22. Macleod MR, O’Collins T, Howells DW, et al. Pooling of animal experimental http://openscience.bmj.com/ 35 36 data reveals influence of study design and publication bias. Stroke 2004;35:1203- 37 38 1208. 39 40 41 23. Hooijmans C.R., Rovers, M.M., de Vries, R.B. et al. SYRCLE’s risk of bias tool 42 43 for animal studies. BMC Med Res Methodol 2014, 14: 43.

44 on September 27, 2021 by guest. Protected copyright. 45 https://doi.org/10.1186/1471-2288-14-43. 46 47 48 24. Nikolakopoulou A; Higgins J.P.T; Papakonstantinou T; Chaimani A; Del Giovane 49 50 C; Egger M, et al. CINeMA: An approach for assessing confidence in the results 51 of a network meta-analysis. PLoS Med 2020; 17(4): e1003082. 52 53 54 55 25. Percie du Sert N; Hurst V; Ahluwalia A; Alam S; Avey M.T; Baker M, et al. The 56 57 ARRIVE guidelines 2.0: Updated guidelines for reporting animal research. PLoS 58 Biol 2020; 18(7): e3000410. 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 20 of 22

19 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 26. Mchugh M. L; Interrater Reliability: The Kappa Statistic. Biochem Med 2012; 4 5 22(3):276-82. 6 7 8 27. Ramos-Hryb A. B; Harris C; Aighewi O; Lino-de-Oliveira C. How would 9 10 publication bias distort the estimated effect size of prototypic antidepressants in 11 12 Confidential:the forced swim test? Neuroscience For & Biobehavioral Review Reviews Only 2018; 92: 192-194. 13 14 15 28. R Development Core Team. An Introduction to R. R Foundation for Statistical 16 17 Computing, Vienna, Austria 2008. ISBN 3-900051-12-7 available in 18 19 https://www.r-project.org/about.html. 20 21 22 29. Thomas et al. Living systematic reviews: 2. Combining human and machine 23 24 effort. Journal of Clinical Epidemiology 2017; 91:31-37. 25 26 27 30. Elliott et al. Living systematic review: 1. Introduction-the why, what, when, and 28 29 how. Journal of Clinical Epidemiology 2017; 91:23-30. 30 31 32 33 34 35 http://openscience.bmj.com/ 36 37 38 39 40 41 42 43

44 on September 27, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from

Page 21 of 22 BMJ Open Science

1 2 3 4 5 6 7 8 9 10 11 Confidential: For Review Only 12 13 14 15 16 Figure 1 - Timeline of a systematic review and meta-analysis. Note: Icons in the illustration are available in 17 flaticon.com 18 19 160x44mm (300 x 300 DPI) 20 21 22 23 24 25 26 27 28 29 30 http://openscience.bmj.com/ 31 32 33 34 35 36 37 38 39 40 on September 27, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmjos BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from

BMJ Open Science Page 22 of 22

1 2 3 4 5 6 7 8 9 10 11 Confidential: For Review Only 12 13 14 15 16 17 18 19 20 21 22 23 Figure 2 - Adult hippocampal neurogenesis. Note: Icons in the illustration are available in flaticon.com 24 25 160x80mm (300 x 300 DPI) 26 27 28 29 30 http://openscience.bmj.com/ 31 32 33 34 35 36 37 38 39 40 on September 27, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmjos BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from

Page 23 of 22 BMJ Open Science

1 2 3 4 5 6 7 8 9 10 11 Confidential: For Review Only 12 13 14 15 16 17 18 19 20 21 Figure 3 - Types of primary studies included in every systematic review according to population, 22 intervention, comparison and outcome. Note: Icons in the illustration are available in flaticon.com 23 24 25 26 27 28 29 30 http://openscience.bmj.com/ 31 32 33 34 35 36 37 38 39 40 on September 27, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmjos BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from

BMJ Open Science

Confidential: For Review Only A protocol for systematic review and meta-analysis of the evidence linking hippocampal neurogenesis to the effects of antidepressants on mood and behaviour

Journal: BMJ Open Science

Manuscript ID bmjos-2020-100077.R2

Article Type: Protocol

Date Submitted by the 25-Nov-2020 Author:

Complete List of Authors: Bolzan, Juliana; Universidade Federal de Santa Catarina, Departamento de Ciências Fisiológica; Universidade Federal de Santa Catarina, Pós- Graduação em Farmacologia Lino de Oliveira, Cilene; Universidade Federal de Santa Catarina, Departamento de Ciências Fisiológicas; Universidade Federal de Santa Catarina, Pós-Graduação em Farmacologia

Keywords: hippocampus, neurogenesis, antidepressants, effect size, animal models http://openscience.bmj.com/

on September 27, 2021 by guest. Protected copyright.

https://mc.manuscriptcentral.com/bmjos Page 1 of 23 BMJ Open Science BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 4 5 6 7 8 9 I, the Submitting Author has the right to grant and does grant on behalf of all authors of the Work (as defined 10 in the below author licence), an exclusive licence and/or a non-exclusive licence for contributions from authors 11 who are: i) UK Crown employees; ii) where BMJ has agreed a CC-BY licence shall apply, and/or iii) in accordance 12 with the termsConfidential: applicable for US Federal Government For officers Review or employees actingOnly as part of their official 13 duties; on a worldwide, perpetual, irrevocable, royalty-free basis to BMJ Publishing Group Ltd (“BMJ”) its 14 licensees and where the relevant Journal is co-owned by BMJ to the co-owners of the Journal, to publish the 15 Work in this journal and any other BMJ products and to exploit all rights, as set out in our licence. 16 17 The Submitting Author accepts and understands that any supply made under these terms is made by BMJ to 18 the Submitting Author unless you are acting as an employee on behalf of your employer or a postgraduate 19 student of an affiliated institution which is paying any applicable article publishing charge (“APC”) for Open 20 Access articles. Where the Submitting Author wishes to make the Work available on an Open Access basis (and 21 intends to pay the relevant APC), the terms of reuse of such Open Access shall be governed by a Creative 22 Commons licence – details of these licences and which Creative Commons licence will apply to this Work are set 23 out in our licence referred to above. 24 25 Other than as permitted in any relevant BMJ Author’s Self Archiving Policies, I confirm this Work has not been 26 accepted for publication elsewhere, is not being considered for publication elsewhere and does not duplicate 27 material already published. I confirm all authors consent to publication of this Work and authorise the granting 28 of this licence. 29 30 31 32 33 34 35 http://openscience.bmj.com/ 36 37 38 39 40 41 42 43

44 on September 27, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 2 of 23

1 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 Title: A protocol for systematic review and meta-analysis of the evidence linking 4 5 6 hippocampal neurogenesis to the effects of antidepressants on mood and behaviour 7 8 Authors: Juliana A Bolzan 1,2, Cilene Lino de Oliveira1,2. 9 10 11 Affiliations: 12 Confidential: For Review Only 13 14 1- Department of Physiological Sciences, Center of Biological Sciences, Federal 15 16 University of Santa Catarina – UFSC, Campus Trindade, 88040-900, Florianópolis - SC, 17 18 Brazil. 19 20 21 2- Post graduation in Pharmacology, Center of Biological Sciences, Federal University 22 23 24 of Santa Catarina – UFSC, Campus Trindade, 88040-900, Florianópolis - SC, Brazil. 25 26 Address to correspondence: 27 28 29 Cilene Lino de Oliveira 30 31 32 LABORATORY OF BEHAVIORAL NEUROBIOLOGY (LABNEC-CFS-UFSC)- 33 34

Departmento de Ciências Fisiológicas, Centro de Ciências Biológicas, Universidade http://openscience.bmj.com/ 35 36 37 Federal de Santa Catarina. Campus Universitário, Trindade, CEP: 88040-900, 38 39 Florianópolis, SC – Brazil, Phone: +55 48 3721-7085, E-mail: [email protected], 40 41 Orcid: orcid.org/0000-0002-0627-530X 42 43

44 Running title: Hippocampal neurogenesis and antidepressants. on September 27, 2021 by guest. Protected copyright. 45 46 Word counting (except the title page, abstract, references, figures, tables or 47 48 49 acknowledgements): 2.915 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos Page 3 of 23 BMJ Open Science

2 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 Authors’ contributions: 4 5 6 7 CLO: selected research theme; performed preliminary studies; wrote the preliminary 8 9 protocol; wrote the manuscript; approved final version of the manuscript. 10 11 JAB: performed preliminary studies; wrote the preliminary protocol; wrote the 12 Confidential: For Review Only 13 14 manuscript; approved the final version of the manuscript. 15 16 Funding: This study was financed in part by the Coordenação de Aperfeiçoamento de 17 18 Pessoal de Nível Superior – Brasil (CAPES) – Finance Code 001. Juliana A Bolzan is a 19 20 recipiente of a fellowship from Conselho Nacional de Desenvolvimento Científico e 21 22 23 Tecnológico, Brazil (131247/2019-0). 24 25 Competing interests: Senior author (Cilene Lino de Oliveira) is an Associate Editor of 26 27 BMJ Open Science. Other than that, no conflict of interest is declared. 28 29 30 Open data: open data are available at (https://osf.io/5z8us/) 31 32 33 34 35 http://openscience.bmj.com/ 36 37 38 39 40 41 42 43

44 on September 27, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 4 of 23

3 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 ABSTRACT 4 5 6 Objective: Studies in rodents associated the deficits of adult hippocampal neurogenesis 7 8 with behavioural anomalies which may be reversed by antidepressant treatments. A 9 10 previous systematic review (SR) and meta-analysis (MA) indicated a hierarchy within the 11 12 pro neurogenicConfidential: effects of different antidepressants For Review in naive rodents. Only The present review 13 14 15 aims to evaluate a more comprehensive sample of studies investigating the links between 16 17 the effects of different antidepressants and adult hippocampal neurogenesis. Search 18 19 strategy, Screening annotation, Data management: Protocols were planned following 20 21 22 PRISMA-P guidelines. Searches in Embase, Medline, Scopus and Web of Science 23 24 followed by screening with inclusion/exclusion criteria will provide relevant publications. 25 26 First SR will summarize the effects of antidepressants on adult hippocampal neurogenesis 27 28 on different laboratory rodents. Second SR will summarize the correlations between 29 30 31 neurogenic and behavioural effects of antidepressants while the third will focus on cause- 32 33 effect relationships between them. If feasible, data will be analyzed by pairwise or 34 http://openscience.bmj.com/ 35 network random-effect or multivariate MA to determine the direction, magnitude, 36 37 38 significance and heterogeneity (I-squared statistics) of the estimated effect sizes on global 39 40 or subgroup levels. Funnel plotting, Egger regression, 'trim and fill' will be used to 41 42 estimate the risk of publication bias. Quality assessment of the included publications will 43

44 on September 27, 2021 by guest. Protected copyright. 45 be performed by applying adapted CAMARADES, Syrcles' RoB or CINeMA tools. 46 47 Reporting: Find a preliminary version of this protocol at the Open Science Framework 48 49 (https://osf.io/gmsvd/). Data extraction has already started. Results shall be published in 50 51 a peer-reviewed journal. Due to the continuous production in the field, the 52 53 54 implementation of a "Living Systematic Review" is intended. 55 56 Keywords: hippocampus, neurogenesis, antidepressants, effect size, heterogeneity, 57 58 animal models. 59 60

https://mc.manuscriptcentral.com/bmjos Page 5 of 23 BMJ Open Science

4 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 STRENGTHS OF THIS STUDY 4 5  6 A preliminary version of this protocol is registered at Open Science Framework. 7 8  Search and screening cover an extensive range of publications. 9 10  Two independent reviewers will perform data extraction. 11 12  Confidential:Quality of studies will be assessed For with Reviewthree different tools. Only 13 14 15  Meta-analyses are planned to include indirect comparisons and non-dependent ES. 16 17 LIMITATIONS OF THIS STUDY 18 19  The protocol is not registered at PROSPERO because data extraction already 20 21 22 began. 23 24  Reviewers will be not blind to the bibliographic information of primary studies. 25 26  A single reviewer performed search and screening, a second one double-checked. 27 28  Correlation between subgroups of studies may not be discarded. 29 30 31 32 33 34 35 http://openscience.bmj.com/ 36 37 38 39 40 41 42 43

44 on September 27, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 6 of 23

5 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 INTRODUCTION 4 5 6 7 The consumption of antidepressant medications used in psychiatric treatments 8 9 doubled according to the Organization of Economic Cooperation and Development 10 11 (OECD) between the years 2000 and 2015 [1]. Despite a large number of patients under 12 Confidential: For Review Only 13 treatment, the mechanisms underlying therapeutic effects of antidepressants remain 14 15 16 elusive. Behavioural studies indicate that the therapeutic effects of antidepressants 17 18 depend on the reversal of neurochemical deficiencies [2]. For example, earlier studies 19 20 revealed that most antidepressants inhibit the reuptake of monoamines, promoting high 21 22 [3] 23 extracellular concentrations in the central nervous system . Moreover, antagonism of 24 25 monoaminergic transmission provoked sudden decay of the effectiveness of 26 27 antidepressants in patients [4, 5]. Although durable, the relationships between monoamines, 28 29 depression and antidepressants may not explain all aspects of antidepressant therapy. For 30 31 32 example, the delay between the onset of treatment and remission of symptoms or the 33 34 existence of depressions resistant to prototype antidepressants suggests there are more for 35 http://openscience.bmj.com/ 36 depression than monoamines. 37 38 39 Animal models allow for discoveries generating hypothesis and theories on 40 41 depression and antidepressant efficacy beyond monoamines. Neurogenic theory of 42 43 depression gained more attention in the 2000s when a series of papers reported that 44 on September 27, 2021 by guest. Protected copyright. 45 46 chronic antidepressant treatment increased the number of newborn neurons in the 47 48 hippocampus of adult rodents [6,7,8] suggesting a pro neurogenic effect of the treatment. 49 50 Stressful stimuli provoked hippocampal cell loss and behavioural sequelae prevented by 51 52 chronic antidepressant treatment in rats [9] indicating a positive correlation between pro 53 54 55 neurogenic and behavioural effects of the compounds. Furthermore, ablation of 56 57 hippocampal neurogenesis reverted the effects of antidepressants on the behaviour of 58 59 adult mice [7,8,10] suggesting a cause-effect relationship between pro neurogenic actions 60

https://mc.manuscriptcentral.com/bmjos Page 7 of 23 BMJ Open Science

6 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 of antidepressants and their effects on behaviour. Altogether, these data engendered an 4 5 6 explanation considering hippocampal neurogenesis as the underlying mechanism for the 7 8 efficacy of antidepressants on treating mood and behavioural disorders. 9 10 Since the neurogenesis is a time-consuming process, the neurogenic theory offers 11 12 a plausibleConfidential: explanation to delay between For the onsetReview of antidepressant Only treatment and the 13 14 15 therapeutic effect. Neurogenic theory inspired a long series of studies aiming to 16 17 understand details of the process which could provide a variety of new targets for the 18 19 development of antidepressants. Over time the findings diverging of the earliest 20 21 22 publications emerged indicating that some effects of antidepressants could be related to 23 24 hippocampal neurogenesis while others would be unrelated to it [11,12]. For example, the 25 26 effects of chronic antidepressants on behavioural tests such as novelty suppressed feeding 27 28 seem to correlate positively and depend on hippocampal neurogenesis [7]. In contrast, the 29 30 31 effects of antidepressants on the forced swimming test seem unrelated to it [11]. A pioneer 32 33 systematic review and meta-analysis of studies investigating effects of antidepressants in 34 http://openscience.bmj.com/ 35 the hippocampus of adult laboratory rodents suggested that some compounds increased 36 37 [13] 38 the neurogenesis while others did not . Depending on the type of the compound 39 40 (fluoxetine, imipramine, venlafaxine, desvenlafaxine), laboratory species (rat or mice) 41 42 and the type of molecular marker (BrdU, doublecortin, Ki-67), the magnitude of the effect 43 44 [13] on September 27, 2021 by guest. Protected copyright. 45 size (ES) may vary from negligible to very large . The broad range of ES values 46 47 indicates that actions of antidepressants may be more or less related or dependent on 48 49 hippocampal neurogenesis. 50 51 To further explore the concept of the hierarchy of effect sizes within different 52 53 54 antidepressants, present work proposes to analyze the data of hippocampal neurogenesis 55 56 with network meta-analysis. Network meta-analysis may enable indirect comparisons 57 58 between effect sizes [14]. Additionally, the present protocol also plans to include studies 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 8 of 23

7 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 of all times to perform a comprehensive systematic review and meta-analysis, which may 4 5 [13] 6 help to overcome some of the limitations of the previous one . Although enough 7 8 powered to evaluate effect sizes of different antidepressants on different species of 9 10 laboratory rodents, boundaries of dates (five years between 2013-2018) may have reduced 11 12 the amountConfidential: of data available to assess For other sourcesReview of heterogeneity Only [13]. Indeed, the 13 14 15 information was insufficient to investigate the influence of relevant characteristics of the 16 17 population such as sex, age and co-morbidities (stressed, anxious, depressed phenotypes) 18 19 [13]. Scarcity of relevant data also precluded the stratified analysis of the effect sizes of 20 21 [13] 22 the antidepressants by doses, via of administration or efficacy in behavioural testing . 23 24 Considering there is a variety of primary studies investigating the theme, it seems 25 26 not feasible to obtain all relevant information with a single systematic review. In a pilot 27 28 analysis, at least three categories of studies investigating the links between hippocampal 29 30 31 neurogenesis and antidepressants in laboratory rodents were identified in the literature. 32 33 For convenience, these categories were here named as 1- interventional, 2- correlational, 34 http://openscience.bmj.com/ 35 3-functional. Interventional category comprised those studies on the effects of the 36 37 38 intervention (antidepressant) on the different markers of the hippocampal neurogenesis 39 40 (neurogenic outcome), for example [6,7,8]. The correlational category included the studies 41 42 on the effects of the antidepressant (intervention) on a behavioural testing (outcome) and 43 44 [15,16] on September 27, 2021 by guest. Protected copyright. 45 different phases of the hippocampal neurogenesis (outcome), for instance . 46 47 Functional studies were those making ablation of hippocampal neurogenesis 48 49 (intervention), co-treated or not with antidepressants, before behavioural testing 50 51 (outcome), for example [7]. Therefore, three different systematic reviews (SR) were 52 53 54 planned to summarize interventional (SR 1), correlational (SR 2) and functional (SR 3) 55 56 studies. In summary, SR 1 will further extend and update previous meta-analysis [13] while 57 58 SR 2 and SR 3 will explore newer aspects of the subject. 59 60

https://mc.manuscriptcentral.com/bmjos Page 9 of 23 BMJ Open Science

8 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 The “neurogenic outcome”, appraised in SR 1 and 2, often means the combination 4 5 6 of different outcomes (measures of stemness, proliferation, survival, maturation of 7 8 newborn neurons). Besides, every behavioural testing, evaluated in SR 2 and 3, should 9 10 provide a different behavioural outcome (e.g. immobility time in the forced swimming 11 12 test, latencyConfidential: to touch the food in the noveltyFor suppressed Review feeding, Onlyso on). Since random- 13 14 15 effects model assumes independence of the effect sizes, but often information from the 16 17 primary studies did not allow to discard correlations between them, current protocol plan 18 19 to apply multivariate meta-analysis to take non-independence of data into account [17]. 20 21 22 23 METHODS AND ANALYSIS 24 25 26 Overview of the systematic reviews and meta-analyses. 27 28 29 The protocol and the manuscript follow PRISMA-P statement and Syrcle’s 30 31 formulary. However, approval by the ethics committee and the Helsinki Declaration does 32 33 34 not apply. The plan of every systematic review (SR 1, SR 2, SR 3) included the following 35 http://openscience.bmj.com/ 36 steps: 1-research question; 2-choice of search strategy; 3-choice of bibliographic 37 38 databases; 4- choice of inclusion and exclusion criteria; 5- planning of data extraction and 39 40 41 6- planning of the meta-analysis (Figure 1). SR 1, 2 and 3 aim to answer the following 42 43 questions prepared using the PICO tool [18], respectively: 1-do different antidepressants

44 on September 27, 2021 by guest. Protected copyright. 45 increase hippocampal neurogenesis at different degrees in laboratory rodents? 2-are there 46 47 48 correlations between behavioural and neurogenic effects of antidepressants in laboratory 49 50 rodents? 3- is there a causal link between behavioural and neurogenic effects of 51 52 antidepressants in laboratory rodents? 53 54 In the three reviews, the population comprises laboratory rodents independent of 55 56 57 species, strain, age, sex or co-morbidities. Interventions and outcomes vary across SR, as 58 59 follows. The SR 1 includes all studies investigating the effects of antidepressant treatment 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 10 of 23

9 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 (intervention) on adult hippocampal neurogenesis (outcome), independent of the type of 4 5 6 experimental design or association with behavioural testing. SR 2 will summarize the 7 8 correlation between outcomes (hippocampal neurogenesis and behaviour). In those 9 10 studies, the effects of antidepressant treatment (intervention) on adult hippocampal 11 12 neurogenesisConfidential: (outcome) and behaviour For testing Review (outcome) are investigated Only in the same 13 14 15 cohort of laboratory animals. Values of correlation will be extracted from primary studies, 16 17 and when absent, calculated from the extracted data. SR 3 will condense those studies 18 19 investigating the effects of ablating hippocampal neurogenesis (intervention) in 20 21 22 laboratory rodent, with or without co-treatment with an antidepressant, evaluated in the 23 24 selected behaviour testing (outcome). 25 26 Neurogenesis will be assessed through the measures of different molecular 27 28 markers of neurogenesis in the adult hippocampus (BrdU, Ki-67, DCX), arbitrarily 29 30 31 chosen according to previously published SRMA [13]. The different stages of neurogenesis 32 33 (Figure 2) will be estimated by quantification of the following markers on the dentate 34 http://openscience.bmj.com/ 35 gyrus of the hippocampus: 1- Nestin (expressed in stem cells), 2- BrdU (incorporated by 36 37 38 proliferating cell) or Ki-67 (expressed in proliferating cells, or other markers such as 39 40 Tbr2, Mash 1, PCNA), 3- DCX (expressed in immature neurons), 4- BrdU and NeuN 41 42 (expressed in mature neurons or other markers of mature neurons such as Map-2, Fox-3, 43

44 on September 27, 2021 by guest. Protected copyright. 45 Prox-1) co-localized (indicating mature neurons incorporated BrdU and matured in the 46 47 adulthood). Measurements of the markers may be using stereological or non-stereological 48 49 methods. The behavioural outcomes depend on the type of behavioural test reported (e.g., 50 51 time of immobility in forced swimming or tail-suspension test, latency to eat in the 52 53 54 novelty suppressed feeding, so on). Data extract from each SR will be analyzed 55 56 separately. Analytical choices were made taken characteristics of the outcomes into 57 58 account. 59 60

https://mc.manuscriptcentral.com/bmjos Page 11 of 23 BMJ Open Science

10 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 4 5 6 7 Search strategy 8 9 10 The search strategies (see Supplementary Table 1, Supplementary Table 2 and 11 12 SupplementaryConfidential: Table 3 at the Open Science For Framework, Review access link: Only https://osf.io/tajhf/) 13 14 were conceived after pilot studies using combinations of keywords from protocols 15 16 [19,20;21] 17 previously published . Databases for the searches (“Pubmed”, “Scopus”, 18 19 “Embase”, “Web of Science”) were arbitrarily chosen considering their extensive 20 21 collection of international publishers in the field of basic biomedicine and our institutional 22 23 24 access. The keywords for SR 1 (Supplementary Table 1) will be a combination of MeSH 25 26 terms related to the primary outcome (hippocampal neurogenesis, #1-5 ‘neurogenesis’ 27 28 and ‘hippocampus’ and (or ‘BrdU’ or ‘DCX’ or ‘Ki-67’) and intervention 29 30 (antidepressants, #6). The keywords for the SR 2 will be terms related to behavioural 31 32 33 outcomes (#7 Supplementary Table 2) combined with the terms of SR 1 (Supplementary 34 35 Table 1). The search terms exclusive of the SR 2 were chosen by examining the http://openscience.bmj.com/ 36 37 publications obtained in SR 1. The keywords for the SR 3 will be terms related to the 38 39 40 ablation of neurogenesis (#8, Supplementary Table 3) combined with the terms related to 41 42 the behavioural outcomes (#7, Supplementary Table 2) and neurogenic outcomes (#1-5, 43 44 Supplementary Table 1). Terms exclusive of SR 3 were chosen by examining the on September 27, 2021 by guest. Protected copyright. 45 46 [7] 47 references similar to Santarelli et al., 2003 . Publications returned of every search will 48 49 be exported to a reference manager file of EndNote X7. Searches were performed by one 50 51 reviewer (JB) and will be double checked for a second reviewer. 52 53 54 Screening processes: 55 56 57 58 Screening of publications returned from the searches will occur in two rounds, the 59 60 first one analyzing title and abstract and second-round evaluating full text. The exclusion

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 12 of 23

11 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 criteria are reviews, systematic reviews, meta-analyses, duplicates (all SR) and co- 4 5 6 treatment (SR 1 and 2). The first round will exclude reviews, systematic reviews, meta- 7 8 analyses, duplicates. The second round of screening will be required to identify co- 9 10 treatments, exclusion criterion in SR 1 and 2, and apply inclusion criteria for all SRs. 11 12 Confidential:All controlled studies will be For included Review in SRs, regardless Only of randomization, 13 14 15 blinding or study design (Figure 3). There will have no restriction on language or dates. 16 17 Studies on adult rats or mice, regardless of sex, strain or stress background; studies with 18 19 any route, dose and treatment schedule for drug administration; studies measuring any 20 21 22 molecular markers in the dentate gyrus of the hippocampus (see Figure 3) will be eligible. 23 24 In these studies, interventions (antidepressants in SR 1 and 2; ablation of neurogenesis 25 26 with or without co-treatment with an antidepressant, SR 3) will be compared to controls, 27 28 i.e., vehicle (SR 1 and 2) or sham ablation or the absence of the intervention (SR 3). 29 30 31 Although studies with co-treatments will be excluded in SR 1 and 2, if the same study 32 33 includes groups of animals treated with a single antidepressant and the respective control 34 http://openscience.bmj.com/ 35 group, these groups will be included. One reviewer performed the screening process; a 36 37 38 second reviewer will double-check the screening. A first reviewer performed the 39 40 screening process (analyzing title and abstract and evaluating full text); a second reviewer 41 42 will double-check the screening. A third reviewer will solve discrepancies. Systematic 43

44 on September 27, 2021 by guest. Protected copyright. 45 reviewers will not be blinded to authors, date or journal of primary articles. 46 47 48 Quality of study 49 50 51 Quality of every publication will be estimated by the use of an adapted version of 52 53 CAMARADES checklist [22] (Table 1) and RoB Syrcle tools [23] which assess, 54 55 56 respectively, general aspects of experimental design and more specific features of animal 57 58 research. CINeMA tool will be used in the case of the implementation of the network 59 60 meta-analysis [24].

https://mc.manuscriptcentral.com/bmjos Page 13 of 23 BMJ Open Science

12 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 The quality of every study evaluated with adapted CAMARADES checklist will 4 5 6 vary from score zero (minimum) to ten (maximum). Every study may receive a score 7 8 when reporting information required in the checklist anywhere in the text (title, abstract 9 10 or body of the text). This version of the CAMARADES checklist was adapted using items 11 12 6 andConfidential: 9 of ARRIVE 2.0 guidelines [25] For. A score Review at item 9 may be Onlyobtained by reporting 13 14 15 (ARRIVE item 9): "a. What was done, how it was done, and what was used? b. When 16 17 and how often. c. Where (including detail of any acclimatization periods). d. Why 18 19 (provide a rationale for procedures)". A score at item 10 may be obtained by reporting 20 21 22 (ARRIVE item 6): “Clearly define all outcome measures assessed (e.g., cell death, 23 24 molecular markers, or behavioural changes)”. The quality of every study evaluated using 25 26 Syrcle's RoB tool will be performed by answering the ten signalling questions with "yes", 27 28 "unclear" or "no", which will be considered a "low", "unclear" and "high" risk of bias, 29 30 31 respectively [23]. The quality using CINeMA tool [24] considers the following domains: (i) 32 33 within-study bias, (ii) reporting bias, (iii) indirectness, (iv) imprecision, (v) heterogeneity, 34 http://openscience.bmj.com/ 35 and (vi) incoherence. To each domain, threes levels o judgement may be assigned: 1-no 36 37 38 concerns, 2-some concerns, or 3-major concerns. Judgments may be summarised to 39 40 obtain four levels of confidence: 1-very low, 2-low, 3-moderate, or 4-high. Two 41 42 independent reviewers will perform the quality assessment. A third reviewer will solve 43

44 on September 27, 2021 by guest. Protected copyright. 45 discrepancies. 46 47 Table 1 – Adapted CAMARADES' study quality checklist 48 1-peer-reviewed publication. 49 50 2- studies following ARRIVE (or other) guidelines. 51 3- declaration of compliance with animal testing regulations and legislation. 52 4- declaration of interest. 53 54 5- report of the breeding, husbandry conditions and actions to improve animal welfare 55 of the experimental animals (for example, environmental enrichment). 56 6-report of the species, lineage or other identifying characteristics of the experimental 57 58 animals (e.g. types of transgenes or knockouts) or phenotypes of interest (Stressed? 59 Depressed? Anxious? Other?). 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 14 of 23

13 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 7- report of the age, weight or stage of the experimental animals. 4 5 8- report of the sex of the experimental animals. 6 9- report of the methods of behavioural testing and acquisition of the behavioural 7 outcomes. 8 9 10- report of the methods to the detection of molecular markers, including the brand of 10 antibodies and kits, and the acquisition of the molecular outcomes. 11 12 Confidential: For Review Only 13 14 15 16 Outcome extraction 17 18 19 For a qualitative and quantitative summary of the included studies, bibliographic 20 21 information and data on the experimental design, population, intervention, comparison, 22 23 24 and the qualitative and quantitative aspects of the outcomes will be extracted from every 25 26 publication (Table 2). Qualitative data will be obtained from the text or tables. 27 28 Quantitative data will be obtained from the text or tables of publications or using a digital 29 30 ruler to extract them from the charts. When information is entirely unavailable in the 31 32 33 publication, it may request directly by contact with the authors by e-mail of the 34 35 corresponding author. In the absence of a response, within 30 days, the study will be http://openscience.bmj.com/ 36 37 excluded from the analysis. Two independent reviewers will extract data. A third reviewer 38 39 40 will solve discrepancies. Every independent reviewer will input qualitative and 41 42 quantitative data extracted from publications in an Excel sheet structured with a study per 43 44 row and a variable per column. Agreement between reviewers will be assessed by using on September 27, 2021 by guest. Protected copyright. 45 46 47 “comparing columns” tool of the Excel (qualitative data) or by calculating Cohen Kappa 48 49 index (quantitative data). Discrepancies above 20 %, i.e., a minimum 80% of concordance 50 51 considered substantial [26], will be conciliated by the third reviewer. 52 53 Table 2 − Data extracted from each publication 54 55 Bibliographic information Authorship, year, journal 56 57 Experimental design See items in the adapted CAMARADES 58 checklist (Table 1) and RoB Syrcle tool 59 [23] 60

https://mc.manuscriptcentral.com/bmjos Page 15 of 23 BMJ Open Science

14 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 Population Species, strain, sex, age, the timing of 4 5 euthanasia, phenotype of experimental 6 rodents 7 Intervention SR 1-3 (antidepressants): Type, 8 mechanism of action, dose, via, number of 9 10 injections per day. 11 SR 3 (ablation of neurogenesis): the 12 Confidential: Formethod Review employed to Only the ablation 13 Comparison SR 1,2: Type of vehicle. 14 SR 3: the method employed to the sham 15 16 ablation. 17 Outcome (qualitative) SR 1,2: Type of marker, antibodies used, 18 the protocol for detection, protocols for 19 BrdU injection (survival or proliferation), 20 protocols of measuring (stereological or 21 non-stereological). SR 2,3: type, metrics 22 23 and method of acquisition (manual or 24 automatic) of the behavioural outcome. 25 Outcome (quantitative) All SRs (per outcome, for experimental 26 and control groups): mean, standard 27 28 deviation or standard error, sample size 29 (when available, effect sizes with 95% 30 confidence interval), p values or the 31 32 values of F or t, and correlation indexes 33 (SR 2). 34 http://openscience.bmj.com/ 35 Meta-analysis 36 37 38 39 Quantitative data extracted from publications (Table 2) will be used in the meta- 40 41 analysis. A similar protocol of the MA was previously published in the following studies 42 43 [13] [27]. Data of each SR will be analyzed independently. For every study included in each

44 on September 27, 2021 by guest. Protected copyright. 45 SR, an ES will be calculated by using a Normalized Mean Difference and expressed 46 47 48 Hedges’g with the upper and lower 95% confidence limits. For each SR, a global estimate 49 50 of ES will include all laboratory animals, all antidepressants and all markers for 51 52 neurogenesis. In the case of independent outcomes, random-effect models will be used 53 54 55 for the meta-analyses, and heterogeneity will be assessed by using the I-squared statistic. 56 57 Since the random-effects model assumes independence of the ES, when the primary 58 59 studies do not allow to discard correlations between outcomes, multivariate meta-analysis 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 16 of 23

15 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 will be performed. Multivariate meta-analysis may be used to estimate ES and 4 5 [17] 6 heterogeneity taking non-independence of data into account . 7 8 Stratified meta-analyses into the subgroups 1-population, 2-intervention and 3- 9 10 11 outcome are planned (Table 3). Meta-regressions and, if necessary stratified, will be 12 Confidential: For Review Only 13 performed to evaluate doses of the antidepressants. For comparisons between ES of 14 15 different antidepressants, network meta-analyses are planned for data of each SR [14]. 16 17 18 Funnel Plotting, Egger regression, trim and fill, and p-Curve will be used to estimate the 19 20 risk of publication bias. Calculations will be performed using R Studio [28]. 21 22 23 Table 3 – Subgroup meta-analysis 24 25 Population Species, strain, sex, age, phenotype, stress 26 condition and behavioural testing. 27 Intervention Type, mechanism of action, dose, via, 28 29 number of injections per day. 30 Outcome SR 1,2: Type of marker, antibodies used, 31 the protocol for detection, protocols for 32 BrdU injection (survival or proliferation), 33 protocols of measuring (stereological or 34

non-stereological). SR 2,3: type, metrics http://openscience.bmj.com/ 35 36 and method of acquisition (manual or 37 automatic) of the behavioural outcome. 38 39 Prospects: 40 41 42 The results of SR 1, SR 2 and SR 3 and the respective MA will be published 43 44 independently in a peer-reviewed scientific journal. Implementation of a Living on September 27, 2021 by guest. Protected copyright. 45 46 47 Systematic Review (LSR) in this protocol may be useful to keep the systematic reviews 48 49 continuously updated. That is, relevant studies will be incorporated into the previously 50 51 structured database of a formal systematic review as soon as they become available [29][30]. 52 53 54 Tools in bibliographic platforms such as PubMed offer periodical reports on a systematic 55 56 review registered in the system. The LSR format is particularly important when a research 57 58 field generates new data frequently. We expect that researchers interested in 59 60

https://mc.manuscriptcentral.com/bmjos Page 17 of 23 BMJ Open Science

16 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 antidepressant research, especially those worried with mechanisms of action, may benefit 4 5 6 from having updated information on this subject to plan new studies or make new 7 8 synthesis to improve theoretical models in the field. 9 10 11 12 Confidential: For Review Only 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 http://openscience.bmj.com/ 36 37 38 39 40 41 42 43

44 on September 27, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 18 of 23

17 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 REFERENCES 4 5 6 7 1. Organization for Economic Cooperation and Development. Health at a 8 Glance 2015: OECD Indicators, Publishing, Paris, 2015. 9 10 11 12 2.Confidential: Leonard, B. Clinical implications For of mechanismsReview of action Only of antidepressants. 13 Advances in Psychiatric Treatment 2000;6(3): 178−86. 14 15 16 17 3. Liu B. From Serotonin to Neuroplasticity: Evolvement of Theories for Major 18 19 Depressive Disorder. Front Cell Neuroscience 2017;28(11):305. 20 21 22 4. Delgado P.L; D.S. Charney; LH Price, G.K. Aghajanian, H. Landis, G.R. 23 24 Heninger L.H. Serotonin function and the mechanism of antidepressant action: 25 Reversal of antidepressant-induced remission by rapid depletion of plasma 26 27 tryptophan Archives General Psychiatry 1990;47:411-18. 28 29 30 31 5. Delgado P.L; Delgado, H.M; Miller, R.M; Salomon, J; Licinio, A.J; Gelenberg, 32 D.S. Monoamines and the mechanism of antidepressant action: Effects of 33 34 catecholamine depletion on mood in patients treated with antidepressants 35 http://openscience.bmj.com/ 36 Psychopharmacology Bulletin 1993;29:389-96. 37 38 39 6. Malberg J.E; Eisch A.J; Nestler E.J; Duman R.S. Chronic antidepressant treatment 40 41 increases neurogenesis in adult rat hippocampus. Journal of Neurosciences 2000; 42 43 20:9104 –10.

44 on September 27, 2021 by guest. Protected copyright. 45 46 7. Santarelli L; Saxe M; Gross C; Surget A; Battaglia F; Dulawa S; Weisstaub N; 47 48 Lee J; Duman R; Arancio O; Belzung C; Hen R. Requirement of hippocampal 49 50 neurogenesis for the behavioral effects of antidepressants. Science 51 2003;301(5634):805-9. 52 53 54 55 8. Miller B.R; Hen R. The current state of the neurogenic theory of depression and 56 57 anxiety. Curr Opin Neurobiol 2015:51–58. 58 59 60

https://mc.manuscriptcentral.com/bmjos Page 19 of 23 BMJ Open Science

18 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 9. Eliwa H; Belzung C; Surget A. Adult hippocampal neurogenesis: Is it the alpha 4 5 and omega of antidepressant action? Biochemical Pharmacology 2017; 141:86 – 6 7 99. 8 9 10 10. Surget A; Tanti A; Leonardo E.D; Laugeray A; Rainer Q et al. Antidepressants 11 12 Confidential:recruit new neurons to improve For stress responseReview regulation. Only Molecular Psychiatry 13 14 2011; 16:1177 – 88. 15 16 17 11. David D.J; Samuels B.A; Rainer Q; Wang J.W; Marsteller D; Mendez I et al. 18 19 Neurogenesis-dependent and -independent effects of fluoxetine in an animal 20 model of anxiety/depression. Neuron 2009; 62:479 – 93. 21 22 23 24 12. Nollet P.M; Tanti A; Girault V; Belzung C; Lema S. Neurogenesis-independent 25 26 antidepressant-like effects on behavior and stress axis response of a dual orexin 27 receptor antagonist in a rodent model of depression. Neuropsychopharmacology 28 29 2012; 37:2210-21. 30 31 32 33 13. Lino de Oliveira C; Bolzan J.A; Surget A; Belzung C. Do antidepressants promote 34

neurogenesis in adult hippocampus? A systematic review and meta-analysis on http://openscience.bmj.com/ 35 36 naive rodents. Pharmacology & Therapeutics 2020 (Available online) 37 38 10.1016/j.pharmthera.2020.107515. 39 40 41 14. Furukawa T.A; Salanti G; Atkinson L.Z, et al. Comparative efficacy and 42 43 acceptability of first-generation and second-generation antidepressants in the

44 on September 27, 2021 by guest. Protected copyright. 45 acute treatment of major depression: protocol for a network meta-analysis BMJ 46 Open 2016;6: e010919. 47 48 49 50 15. Malberg J.E; Duman R.S; Cell proliferation in the adult hippocampus is decreased 51 by inescapable stress: reversal by fluoxetine treatment. 52 53 Neuropsychopharmacology 2003;28(9):1562-71. 54 55 56 57 16. Ho NF, Hooker JM; Sahay A; Holt D.J; Roffman J.L. In vivo imaging of adult 58 human hippocampal neurogenesis: Progress, pitfalls and promise. Molecular 59 60 Psychiatry 2013; 18:404-16.

https://mc.manuscriptcentral.com/bmjos BMJ Open Science Page 20 of 23

19 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 4 5 17. Van den Noortgate W; López-López J.A; Marín-Martínez F, et al. Meta-analysis 6 7 of multiple outcomes: a multilevel approach. Behav Res 2015; 47:1274-1294. 8 9 10 18. Pico (Systematic Reviews - Research Guide), available online: link: 11 12 Confidential:https://libguides.murdoch.edu.au/systematic/PICO. For Review Only 13 14 15 19. Ramos-Hryb A.B; Bahor Z; McCann S et al. Protocol for a systematic review and 16 17 meta-analysis of data from preclinical studies employing forced swimming test: 18 19 an update. BMJ Open Science 2019;3: e 000035. doi:10.1136/ bmjos-2017- 20 000043. 21 22 23 24 20. McCann S.K, Irvine C, Mead G.E, et al. Efficacy of antidepressants in animal 25 26 models of ischemic stroke: a systematic review and meta-analysis. Stroke 2014; 27 45:3055–63. 28 29 30 31 21. Lazic S.E; Fuss J; Gass P. Quantifying the Behavioural Relevance of 32 33 Hippocampal Neurogenesis. PLoS ONE 2014; 9:11. 34

22. Macleod MR, O’Collins T, Howells DW, et al. Pooling of animal experimental http://openscience.bmj.com/ 35 36 data reveals influence of study design and publication bias. Stroke 2004;35:1203- 37 38 1208. 39 40 41 23. Hooijmans C.R., Rovers, M.M., de Vries, R.B. et al. SYRCLE’s risk of bias tool 42 43 for animal studies. BMC Med Res Methodol 2014, 14: 43.

44 on September 27, 2021 by guest. Protected copyright. 45 https://doi.org/10.1186/1471-2288-14-43. 46 47 48 24. Nikolakopoulou A; Higgins J.P.T; Papakonstantinou T; Chaimani A; Del Giovane 49 50 C; Egger M, et al. CINeMA: An approach for assessing confidence in the results 51 of a network meta-analysis. PLoS Med 2020; 17(4): e1003082. 52 53 54 55 25. Percie du Sert N; Hurst V; Ahluwalia A; Alam S; Avey M.T; Baker M, et al. The 56 57 ARRIVE guidelines 2.0: Updated guidelines for reporting animal research. PLoS 58 Biol 2020; 18(7): e3000410. 59 60

https://mc.manuscriptcentral.com/bmjos Page 21 of 23 BMJ Open Science

20 BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from 1 2 3 26. Mchugh M. L; Interrater Reliability: The Kappa Statistic. Biochem Med 2012; 4 5 22(3):276-82. 6 7 8 27. Ramos-Hryb A. B; Harris C; Aighewi O; Lino-de-Oliveira C. How would 9 10 publication bias distort the estimated effect size of prototypic antidepressants in 11 12 Confidential:the forced swim test? Neuroscience For & Biobehavioral Review Reviews Only 2018; 92: 192-194. 13 14 15 28. R Development Core Team. An Introduction to R. R Foundation for Statistical 16 17 Computing, Vienna, Austria 2008. ISBN 3-900051-12-7 available in 18 19 https://www.r-project.org/about.html. 20 21 22 29. Thomas et al. Living systematic reviews: 2. Combining human and machine 23 24 effort. Journal of Clinical Epidemiology 2017; 91:31-37. 25 26 27 30. Elliott et al. Living systematic review: 1. Introduction-the why, what, when, and 28 29 how. Journal of Clinical Epidemiology 2017; 91:23-30. 30 31 32 33 34 35 http://openscience.bmj.com/ 36 37 38 39 40 41 42 43

44 on September 27, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/bmjos BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from

BMJ Open Science Page 22 of 23

1 2 3 4 5 6 7 8 9 10 11 Confidential: For Review Only 12 13 14 15 16 Figure 1 - Timeline of a systematic review and meta-analysis. Note: Icons in the illustration are available in 17 flaticon.com 18 19 160x44mm (300 x 300 DPI) 20 21 22 23 24 25 26 27 28 29 30 http://openscience.bmj.com/ 31 32 33 34 35 36 37 38 39 40 on September 27, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmjos BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from

Page 23 of 23 BMJ Open Science

1 2 3 4 5 6 7 8 9 10 11 Confidential: For Review Only 12 13 14 15 16 17 18 19 20 21 22 23 Figure 2 - Adult hippocampal neurogenesis. Note: Icons in the illustration are available in flaticon.com 24 25 160x80mm (300 x 300 DPI) 26 27 28 29 30 http://openscience.bmj.com/ 31 32 33 34 35 36 37 38 39 40 on September 27, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmjos BMJ Open Science: first published as 10.1136/bmjos-2020-100077 on 4 March 2021. Downloaded from

BMJ Open Science Page 24 of 23

1 2 3 4 5 6 7 8 9 10 11 Confidential: For Review Only 12 13 14 15 16 17 18 19 20 21 Figure 3 - Types of primary studies included in every systematic review according to population, 22 intervention, comparison and outcome. Note: Icons in the illustration are available in flaticon.com 23 24 25 26 27 28 29 30 http://openscience.bmj.com/ 31 32 33 34 35 36 37 38 39 40 on September 27, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmjos