The in vitro effects of Zika on human brain organoids: a systematic review

Author : Bayu SutarjonoAffiliation : Saba University School of MedicineArticle word count : 7,053 wordsHypothesis : ZIKV causes congenital microcephaly in the human brain organoid model.

1

Abstract

Background: The rapid spread of the Zika virus (ZIKV) and its devastating effects in the developing fetus has resulted in a global health emergency. The emergence of human brain organoids represents a novel method to investigate early brain development coinciding with the critical period for ZIKV-infection vulnerability of the fetus.

Hypothesis: ZIKV causes congenital microcephaly, as investigated using human brain organoid models.

Methods: Studies were gathered by searching MEDLINE/Pubmed, LILACS, and LiSSa of the effects of ZIKV infection on human brain organoids. The aim of the systematic review was to verify the in vitro effects of ZIKV on the human brain organoid model.

Results: From 146 identified papers, 13 articles were selected for review. In summary, this review found that ZIKV of African, Latin American, and Asian lineages caused productive replication after 72 hours, preferentially infected neural progenitor cells over mature neurons, reduced both cell populations, and caused premature differentiation. Limited data involving only African and Latin American lineages showed a reduction in populations of proliferating cells and intermediate cells, and overall decreased viability. Furthermore, all three lineages caused heightened apoptosis and reduced organoid size.

Conclusion: This systematic review strengthened the hypothesis that ZIKV causes congenital microcephaly, as investigated in the human brain organoid model.

Word count : 205Keywords : Zika, microcephaly, organoid, neural progenitor cells, apoptosis

Ultra-mini abstract

A systematic review identified 13 articles that strengthen the hypothesis that Zika virus causes microcephaly. These studies, using the novel human brain organoid model, revealed that Zika of African, Latin American, and Asian lineages caused productive replication, infected neural progenitor cells over neurons and reduced both populations, induced apoptosis, and reduced organoid size.

Word count : 53

2

Introduction

Rationale

The rapid spread of the Zika virus (ZIKV) and its devastating effects in the developing fetus has

resulted in a global health emergency. ZIKV was introduced into Brazil from the Pacific Islands before

spreading rapidly throughout the Americas1. It causes a self-limiting flu-like febrile illness that resolves

within days, and is symptomatic in only 20% of infected individuals2. However, there is mounting

evidence that a causal relationship exists between maternal ZIKV infection and fetal brain anomalies,

such as cerebral calcification, ventriculomegaly, malformations of the corpus callosum or cortex, or

more strikingly, microcephaly3. One study discovered that 42% of live-born infants who had been

exposed to ZIKV in utero had abnormal findings in the first month of life2.

Currently, there are three main lineages of ZIKV circulating the world: African, Asian, and Latin

American. Only strains from Brazil, Puerto Rica, and French Polynesia cause microcephaly in human

fetuses, which are descendants of the Asian strain1. Yet, the Asian strain does not cause neurological

abnormalities despite proving the contrary among in vitro studies. The African strain, isolated from

Uganda in East Africa and has no connection with the contemporary strains in Latin America, also does

not cause microcephaly4. Thus, it is possible that the pathogenic mechanisms used by the various

lineages are different from one another.

The study of early human brain development is challenging due to ethical and practical issues.

Live ZIKV-infected human fetal tissues are inaccessible and postmortem tissues vary in quality and

genetic makeup, while clinical studies alone cannot provide sufficient insight for the pathogenesis of

ZIKV. Yet, there are limitations to the use of animal models as an alternative. For example, mice are

resistant to ZIKV infection. In order to sustain high levels of ZIKV infection in the brain, mouse models

must remove key components of its antiviral immunity7. Furthermore, mice also lack the outer

subventricular zone, considered pivotal to the evolutionary increase in human cortex size and

complexity8,9. Cortical development in humans is different from rodents, from longer cell cycle length,

longer duration of cortical neurogenesis, bigger brain surface, and complex gyrencephalic instead of

lissencephalic configuration10. Thus, it may be possible that there are aspects of ZIKV infection that are

unique to the human brain.

Human brain organoids represent a novel method to investigate early human brain

development that coincides with the early first trimester in humans, a critical period where the maternal

3

decidua, fetal placenta, and umbilical cord are especially permissive to ZIKV infection11. Organoids are

3D cell aggregates generated in vitro from human pluripotent stem cells, either induced pluripotent

stem cells or embryonic stem cells12. In comparison to 2D cultures prepared on flat surfaces, 3D culture

models possess cell-cell and cell-matrix interactions, and mimic cellular functions and signaling pathways

similar to in vivo conditions, accurately depicting a more natural environment experienced by cells in a

living organism. There are two types of human brain organoids: cerebral organoids and region-specific

organoids. Due to differences in generation techniques, cerebral organoids model whole brain

development and interactions among different brain regions, while region-specific organoids model

individual brain regions of interest, resulting in more uniform and reproducible tissue13.

In the developing human cerebral neocortex, neuroepithelial stem cells transition into radial glia

cells, also known as neural progenitor cells, and reside in both the ventricular and subventricular zones,

giving rise to intermediate progenitors before transforming as neurons, astrocytes, and

oligodendrocytes9,14. A key contributor to human neocortical growth is the expansion of subventricular

zone progenitors, while defects in this process result in a range of neurological disorders, especially

microcephaly15,16. Recently, it has been shown that ZIKV targets neuroepithelial stem cells and neural

progenitor cells, impairing mitosis and survival15.

Brain organoids contain a ventricular zone, with NESTIN-, SOX2-, PAX6-, or phospho-vimentin-

positive neural progenitor cells at the apical side facing a lumen, and CTIP2- or TUJ1-positive mature

neurons at the basal side spatially distinct from the lumen and ventricular zone18. TBR1- or TBR2-

positive cells are intermediate progenitor cells found prominently in the subventricular zone. Cell

proliferation is measured by Ki67 marker or phosphate-H3, while cell viability or cell death is determined

by EdU labeling. Apoptotic cells are measured using TUNEL or Caspase-3 (CASP3) assay.

The recent development of human brain organoids provides a unique opportunity to investigate

the causal relationship between ZIKV and microcephaly in a controlled setting where size can be

monitored and individual cells can be studied very quickly, characteristics previously unattainable using

traditional 2D culture techniques. Thus, in order to test the hypothesis that ZIKV cause congenital

microcephaly in the novel human brain organoid model, the aim of this systematic review was to assess

the evidence for causality between the varying strains of ZIKV and microcephaly in organoids, and to

describe a preliminary mechanism of ZIKV pathogenesis.

Objectives

4

Using the working hypothesis that ZIKV causes congenital microcephaly in human brain

organoids, this study will systematically review published literature of the effects of ZIKV infection on

human brain organoids and identify which cells are most vulnerable to ZIKV infection. This study will

also determine if different ZIKV strains (i.e. African, Asian, or Latin American) induce similar or different

effects.

Methods

Protocol and registration

The Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) Checklist

was adhered to in this systematic review. Protocol was not registered.

Eligibility criteria

Inclusion criteria. Articles that compared the effects of ZIKV of any lineage (i.e. African, Asian,

or Latin American) on human cerebral or region-specific organoids to mock-infected controls were

selected for the systematic review. There was no discrimination as to the cell lines used to develop the

organoids (i.e. human embryonic stem cells or induced pluripotent stem cells). An adapted PICOS

(population, exposure/intervention, comparison, outcome, study design) was applied with the following

criteria:

Population: Human brain organoids.

Cell line: Human embryonic stem cells (HESC) or induced pluripotent stem cells (IPSC).

Exposure: Infection by ZIKV.

Comparisons: Human brain organoids that were not exposed to ZIKV but received a control

treatment.

Outcome: Organoid size, neuronal cell marker expression, apoptosis, cell viability, and viral

copies.

Study design: In vitro studies designed to assess cellular, biochemical, and/or molecular

mechanisms of ZIKV neurotoxicity.

Exclusion criteria. Studies were excluded if: 1) Organoids that were not of human origin (e.g.

mice, chimp); 2) Organoids were not of brain origin (e.g. prostate, syncytiotrophoblasts); 3) Organoids

5

that differentiated to tumour progeny; 4) Reviews of literature, protocol summaries, editorials, opinions

pieces, conference abstracts, and book chapters; or 5) Clinical studies.

Information sources and search strategies

A criteria search was performed using the following electronic databases: MEDLINE/PubMed,

LiSSa (Littérature Scientifique en Santé), and LILACS (Literatura Latino Americana em Ciências da Saúde).

A comprehensive search strategy of the medical literature across all databases was conducted up to and

including February 1, 2018. The search terms using the word ‘Zika’, the variation ‘ZIKV’, and ‘flavivirus’

were searched in combination with the words ‘organoid’, ‘in vitro’, ‘culture’, ‘3D’ (from 3D culturing),

‘bioreactor’ (a technique in order to develop organoid), and ‘stem cells’, ‘progenitors’, ‘HESC’, or ‘IPSC’

(sources of derivations of the organoids). The search strategy was adapted for each database. The

complete search was as follows:

(Zika OR ZIKV OR Flavivirus) AND (Organoid OR Organoids OR In vitro OR Culture OR Cultures OR

Culturing OR 3D OR Bioreactor OR Bioreactors OR stem cells OR Stem cell OR Progenitor OR

Progenitors OR HESC OR Human embryonic stem cell OR Human embryonic stem cells OR IPSC or

Induced pluripotent stem cell OR Induced pluripotent stem cells).

Study selection. Initial triage of articles were based on whether titles and abstracts met the

inclusion criteria. Articles that matched the initial checks were collected for review. Full-text articles

and any supplementary material were then read, and articles that were not in full agreement with the

inclusion criteria were excluded. A PRISMA flow diagram was produced to indicate the number of

included and excluded studies and the corresponding reasons for exclusion (Fig 1).

Data collection process and data items. Data extraction was gathered from the main research

article or supplementary material. These included name of first author, year of publication, country, and

study design. Variables for which the data were sought included the type of experimental model (i.e.

region-specific development such as cerebral, cortical, forebrain, midbrain, hypothalamus), age of

experimental model, ZIKV strain, protocol of infection including duration of effects and multiplicity of

infection (MOI) or viral dilution, physical changes (i.e. size, growth rate, area) of the organoids following

ZIKV infection, and effects of specific types of cells based on markers used (i.e. SOX2, PAX6, phospho-

vimentin, NESTIN, CTIP2,TBR1, TBR2, Ki67, phosphate-H3, TUJ1, TUNEL, CASP3, and EdU), including cell

death and cell viability. Statistical data were gathered only if measurements were compared to mock

infections (i.e. organoids that were not infected by ZIKV). Two evidence tables were produced that

included a summary of the assays, results, and main conclusion of each study (Tables 1 and 2). A

6

separate table was produced that combined strains according to lineages (i.e. African, Latin American,

and Asian) with the broad effects on viral replication after 72 hours, preferential infectivity with regards

to cell type, and the population outcomes of progenitor cells, intermediate cells, proliferating cells,

mature neurons, apoptosis, viability, and change in organoid size (Table 4).

Risk of bias in individual studies. The National Toxicology Program Office of Health Assessment

and Translation (OHAT) Risk of Bias Rating Tool for Human and Animal Studies by the US Department of

Health and Human Services was used to methodically appraise the quality of evidence of all studies. The

risk of bias in experimental methodology was assessed by answering 8 questions, tailored to

accommodate in vitro studies. Questions were rated on a four-point scale depending on standardized

responses: ‘definitely low’, ‘probably low’, ‘probably high, or definitely high’ risk of bias. For a study to

be accepted for the systematic review and rated as high quality, two key risk of bias questions (Exposure

Characterization and Outcome Assessment) must be rated as ‘definitely low’ or ‘probably low’ risk of

bias and have the majority of other applicable items answered ‘definitely low’ or probably low’ risk of

bias. A study that is rated as ‘definitely high’ or ‘probably high’ risk of bias for these two key elements

and have most other applicable items answered ‘definitely high’ or ‘probably high’ risk of bias will be

rated as low quality and excluded from the review.

Summary measures. Cerebral organoid size, viral copies, cell population, and apoptosis

following ZIKV exposure were the main evaluated outcomes.

Results

Study selection

A total of 146 articles were selected from three electronic databases. 27 articles were selected

based on their relevance to ZIKV and in vitro human brain organoid experimentation. 15 articles

remained following the removal of reviews, protocol summaries, editorials, and opinion pieces. The

article by Nowakowski et al. (2016)17 was removed as it contained only theoretical predictions and no

experimentation involving ZIKV. The article by Zhu et al. (2017)18 was also removed as it involved

experimentation using glioblastoma organoids. The article by Sacramento et al. (2017)19 was included

as, although they did not use mock-infected controls for organoid experiments, mock-infected controls

were used in other non-3D neural stem cell in vitro models. Finally, the article by Yoon et al. (2017)20

7

was also included as they expressed ZIKV protein NS2A in developing organoids via electroporation

mimicking ZIKV exposure. Thus, a total of 13 articles fully complied to the study selection

criteria19,20,21,22,23,24,25,26,27,28,29,30,31. All thirteen articles reported original research based on in vitro

experimentation on human brain organoids using ZIKV. A flow chart detailing the process of

identification, inclusion, and exclusion of studies is shown in Fig 1.

Study characteristics

A summary of the study characteristics is found in Table 1. All studies were published published

between 2016 and 2017. The majority of studies were conducted in the USA20,22,26,27,28,30,31, while three

studies took place in Brazil19,21,24. Germany23 and China25 each contributed to one study.

Organoid. While the majority of studies used cerebral organoids19,21,22,23,24,26,28, four used

forebrain-specific organoids20,27,30,31 and two used cortical-specific organoids25,28. No study used

organoids that were specific to other regions of the brain (e.g. midbrain, hypothalamus). The age of

organoids at the time of infection was 15-28 days for 53.3% of experiments21,25,26,27,28,19,30,31. Three

experiments24,25,29 used younger organoids (0-14 days) with four20,24,26,28 conducting research on older

organoids (29+ days). Sacramento et al., (2017)19 did not report the age of the organoid used.

Strains. Five different strains were used. The majority of articles20,21,22,23,24,26,27,31 used the strain

MR766 from Uganda, Africa (African strain). Strain FSS13025 from Cambodia in South East Asia (Asian

strain) was used in two articles27,30. The contemporary Latin America strains were isolated from different

origins. Strain PRVABC59, used in three studies25,28,29, originated from Puerto Rico. Although three

studies19,21,23 sourced their strains from patients in Brazil, the strain used by Gabriel et al. (2016)23

originated from the brain of a fetus exhibiting neurological abnormalities. Only one study23 used strain

H/PF/2013 originating from the French Polynesia (included as part of the Asian lineage for data

assessment purposes).

Infection protocol. In terms of infection protocol, five articles exposed the organoids to ZIKV for

2 hours19,24,25,28,29, while more than half of the studies21,22,23,26,27,30,31 required an incubation period of 24

hours or more. Yoon et al. (2017)20 was the sole study to express ZIKV non-structural protein during the

fabrication process. All studies except for Sacramento et al. (2017)19 reported the use of mock-infected

controls for organoid experiments.

Risk of bias

8

The OHAT Risk of Bias Rating Tool for Human and Animal Studies was used to assess the quality

of the studies. All 13 article were rated as high quality in which, as defined by the protocol, all studies

were assessed ‘definitely low risk of bias’ for questions concerning ‘Exposure Characterization’ and

‘Outcome Assessment’, as well as low risk of bias for the majority of remaining questions. However,

perceived biases were present. 11 studies did not blind their research personnel to their respective

groups20,21,22,23,24,25,26,27,29,30,31, indicating a high risk of bias. Six studies had incomplete data that warranted

a probable high risk of bias. In the study by Gabriel et al. (2017)23, one or two out of three ZIKV strains

were omitted in measurements for cortical plate thickness, organoid diameter, and TUJ1-positive cells

within the ventricular zone lumen. As well, values for TUNEL assay were not recorded. Studies by Dang

et al. (2016)22, Li C et al. (2017)25 and Sacramento et al. (2017)19 did not provide the number of organoids

used for some or all experiment despite calculating statistical results. As well, Sacramento et al. (2017)19

was the only study that did not use mock-infected controls for organoid experiments, indicating a

probable high risk of bias for the categories of ‘Randomization of ZIKV Exposure’ and ‘Allocation

Concealment’. Li Y et al. (2017)26 and Qian et al. (2016)27 had neither provided the number of organoids

used for each experiment, nor disclosed the full statistical method. Furthermore, Li Y et al. (2017)26 did

not quantify apoptosis measurements by CASP3 assay, while Qian et al. (2016)27 did not report the

results for all strains in SOX2- and TBR2-positive cell measurements. A table detailing the results of the

OHAT Risk of Bias evaluation is found in Table 3.

Results of individual studies

A summary of assays and results can be found in Table 2. The articles selected for the review

were divided into three groups based on common research goals.

Comparison of ZIKV strains. Articles by Cugola et al. (2016)21, Gabriel et al. (2017)23, and Qian et

al. (2016)27 investigated the effects of different ZIKV strains on human brain organoids. Comparisons

were made among lineages from Africa, Latin America, and/or Asia.

Cugola et al. (2016)21 evaluated the infection of human brain organoids by the Brazilian (ZIKV-

BR) or African (ZIKV-AF) strains of ZIKV, demonstrating that both reduced proliferative zones and

disrupted cortical layers. After 96 hours of ZIKV-BR infection, there was a 67% decrease in cortical plate

thickness of TBR1-positive intermediate cells (p<0.05 ANOVA) and 60% decrease in CTIP2-positive

neuronal cells (p<0.05 ANOVA) in comparison to mock-infected controls. Measurements using ZIKV-AF

were not reported. After 24 hours of ZIKV-AF infection, there were reductions in SOX2- and PAX6-

positive progenitor cells (-25% for both), Ki67-positive proliferative cells (-60%), and TBR1-positive

9

intermediate cells (-25%), with a decrease in CTIP2-positive neuronal cells at 96 hours only (-25%). In

terms of ZIKV-BR infection, although results for PAX6-positive progenitor cells were different with a

decrease after 96 hours (-50%), outcomes after 24 hours exposure were similar for SOX2-positive

progenitor cells (-37.5%), Ki67-positive proliferative cells (-25%), TBR1-positive intermediate cells (-15%),

and CTIP2-positive neuronal cells (-75% after 96 hours). Apoptotic cell assays revealed an increase in

TUNEL-positive cells and CASP3 activity after 24 hours infection by ZIKV-BR (+15% and +500%

respectively) and ZIKV-AF (+5% and +500% respectively). There were no statistical comparisons

between strains or, except for cortical plate thickness, between mock-infected controls.

Using strain isolated from Asia (ZIKV-AS), Latin America (ZIKV-AM), and Africa (ZIKV-AF), Gabriel

et al. (2017)23 stated that all strains efficiently infected neural progenitors in human brain organoids,

causing premature differentiation closely resembling ZIKV-associated microcephaly. However, the

effects of ZIKV-AS and ZIKV-AM were different than that seen with ZIKV-AF. Despite all three strains

showing comparable increases in viral copies after 72 hours, ZIKV-AS- and ZIKV-AM-infected organoids

exhibited decreased thickness in the ventricular zone as early as 5 days post infection (-13% for ZIKV-AS,

p<0.01 one-way ANOVA; -26% for ZIKV-AM, p<0.001 one-way ANOVA), while ZIKV-AF-infected organoids

presented with the complete destruction of the ventricular zone after 11 days of infection (-100%,

p<0.05 one-way ANOVA) in comparison to mock-infected organoids. However, all strains produced a

reduction in organoid diameter after 11 days of infection (-33% for ZIKV-AS and ZIKV-AM, -66% for ZIKV-

AF, p<0.001 one-way ANOVA). All three strains showed similar decrease in phospho-vimentin-positive

progenitor cells as early as 5 days after exposure (-66% for all three, p<0.001 one-way ANOVA). SOX2-

and PAX6-positive progenitor cells in the ventricular zone were preferentially infected by ZIKV-AS &

ZIKV-AM, whereas ZIKV-AF-infected cells were found distant to SOX2-positive cells in the perimeter. The

investigation into cell death revealed elevated apoptosis as early as 2 days post-infection with

heightened impact by ZIKV-AF (+300% for ZIKV-AF; +60% for ZIKV-AS, p<0.001 ANOVA vs. ZIKV-AF; and

+100% for ZIKV-AM, p<0.001 ANOVA vs. ZIKV-AF). Further comparisons among strains showed greater

severity from ZIKV-AF infected Nestin-positive progenitor cells (-10%, p<0.001 two-way ANOVA) and

TUJ1-positive neuronal cells (-5%, p<0.001 two-way ANOVA), with a corresponding elevated TUNEL-

positive apoptotic cells in areas distinct from the ventricular zone (+20%, p<0.001) when compared to

the other strains.

Using brain-region-specific organoids, Qian et al. (2016)27 discovered preferential and productive

infection of neural progenitors with either African (ZIKV-AF) or Asian ZIKV (ZIKV-CA) strains leading to

increased cell death and reduced proliferation, resulting in decreased neuronal cell-layer volume

10

resembling microcephaly. Human forebrain-specific organoids were found to be permissive to ZIKV-AF,

as there was a rise in viral replication between 4 and 14 days of infection proportional to its

concentration (+40% at 0.25x dilution). ZIKV-CA replication was not measured. In comparison to mock-

infected controls, there was a reduction in organoid area (-83%, infected at 14 days old and analyzed 18

days later, p<0.0005 t-test; -67%, infected at 28 days old and analyzed 14 days later, p<0.0005 t-test)

and 75% decrease in ventricular zone area (p<0.005 t-test), with a subsequent 200% increase in lumen

area (p<0.05 t-test) in ZIKV-AF-infected organoids. The level of destruction was proportional to the

infectivity of ZIKV-AF, leading to decreases in neuronal layer (-50% at 0.25x, p<0.005 t-test) and

ventricular zone (-33% at 0.25x, p<0.05 t-test), as well as the reductions of EdU-positive viable cells in

the ventricular zone (-28% at 0.25x, p<0.005 t-test). Consequently, there was increased CASP3 apoptotic

activity proportional to ZIKV-AF concentration (+800% at 0.25x, p<0.0005 t-test), with a correlated 87.5%

reduction of phosphate-H3-positive cell density (p<0.0005 t-test) when compared to mock-infected

controls. 92.5% of ZIKV-AF enveloped-positive cells were found in SOX2-positive progenitor cells

followed by 5% of CTIP2-positive neuronal cells and 2.5% of TBR2-positive intermediate cells.

Experiments with ZIKV-CA were limited to infection characteristics, which showed comparable results to

ZIKV-AF as 90% of ZIKV-CA enveloped-positive cells were found in SOX2-positive progenitor cells

followed by 5% of CTIP2-positive neuronal cells and 5% of TBR2-positive intermediate cells.

ZIKV pathogenesis. Five articles studied the pathogenesis of ZIKV. Garcez et al. (2016)24 was

one of the first to observe that ZIKV targeted human brain cells causing microcephaly in organoids.

Dang et al. (2016)22 highlighted the role that the innate immune receptor TLR3 played in depleting

neural progenitors. Yoon et al. (2017)20 investigated neurogenesis of the cortex and observed that ZIKV

degraded adherens junction proteins. Wells et al. (2016)31 discovered that AXL, thought to be a ZIKV

entry receptor found on neural progenitor cells, was not important for ZIKV infection. Finally, Li Y et al.

(2017)26 reported that ZIKV restricted cortical growth and folding.

Garcez et al. (2016)24 reported that ZIKV targeted human brain cells, reducing their viability and

growth as brain organoids, suggesting that the virus abrogates neurogenesis during human brain

development. They measured a 40% decrease in area in comparison to mock-infected controls (p<0.05

unpaired two-tailed t-test).

Dang et al. (2016)22 concluded that ZIKV efficiently infected organoids and caused a decrease in

overall organoid size that correlated with the kinetics of viral copy number, linking ZIKV-mediated TLR3

activation, an immune response receptor meditator, leading to perturbed cell fate, and a reduction in

organoid volume reminiscent of microcephaly. There was an increase in ZIKV viral copies after only 24

11

hours (+10,000%). After 5 days infection, the organoid volume was half to that of mock-infected

controls (p<0.05 t-test) with reduced NESTIN-positive progenitor cells.

Yoon et al. (2017)20 investigated the effects of systematically introducing individual proteins

encoding ZIKV-NS2 in human forebrain-specific organoids, concluding that it reduced neural progenitor

cell proliferation and caused adherens junction deficits. They measured a 17.5% reduction of EdU-

positive viable cells and PAX6-EdU-positive viable progenitor cells in ZIKV-NS2A-expressed organoids

(p<0.001 t-test) in comparison to controls. They also found a 27.5% increase in PAX6-positive progenitor

cells with abnormal multipolar morphology (p<0.01 t-test), and 20% decrease in Ki67-positive

proliferating cells (p<0.01 t-test) in ZIKV-NS2A expressed organoids.

Wells et al. (2016)29 declared that the genetic ablation of AXL had no effect on ZIKV entry or

ZIKV-mediated cell death in cerebral organoids. They measured similar decreases in area for AXL

wildtype (AXL-WT) and AXL knockout (AXL-KO) ZIKV-infected organoids as early as 72 hours post

infection (-25% for MOI 0.1), with comparable levels of infection of Ki67-positive proliferating cells and

increases in CASP3 apoptotic activity when compared to mock-infected controls.

Li Y et al. (2017)26 discovered that the deletion of the PTEN gene stimulated proliferation,

producing larger and substantially folded cerebral organoids, and that ZIKV impairs cortical growth and

folding perhaps by interacting with PTEN. They found increased apoptosis for 19 day & 30 day

organoids following ZIKV infection, and 75% and 82% decrease in size in respective control and PTEN-

deleted mutant organoids infected by ZIKV in comparison to mock-infected controls (p<0.001 t-test).

However, they only found a decrease in fold density in PTEN mutant organoids only by ZIKV (-94%,

p<0.001 t-test vs. mock-infected controls).

Drug candidates that inhibit ZIKV infection. Studies by Xu et al. (2016)30, Watanabe et al.

(2017)28, and Zhou et al. (2017)31 used organoids to identify small-molecule candidates to combat ZIKV

virus infection. Sacramento et al. (2017)21 investigated the effectiveness of Sofosbuvir, currently used as

a treatment for hepatitis C, while Li C et al. (2017)25 discovered that 25-Hydroxycholesterol, an amplifier

of inflammatory signaling within the brain, provided protection against ZIKV infection.

Testing the effects of a pan-caspase inhibitor, Xu et al. (2016)30 found that Emricasan inhibited

ZIKV-induced increases in CASP3 apoptotic activity and protected three-dimensional organoid cultures.

They measured a 1.75% increase in CASP3-positive cells after ZIKV infection (p<0.001 one-way ANOVA)

in comparison to mock-infected controls.

Watanabe et al. (2017)28 developed an optimized organoid culture method that efficiently and

reliably produced cortical and basal ganglia structures similar to those in the human fetal brain, and that

12

it models the teratogenic effects of ZIKV on the developing brain. They found that ZIKV-infected

organoids had a smaller area (-14%, p<0.001 paired t-test) and perimeter (-20%, p<0.01 paired t-test) in

comparison to mock-infected controls. They also showed that ZIKV predominantly infected SOX2-

positive progenitor cells (70%, of which 70% are infected), followed by TBR2-positive intermediate cells

(20%, of which 70% are infected), and CTIP2-positive neuronal cells (10%, of which 15% are infected).

There was increased cell death (p<0.001) in ZIKV-infected organoids with 20% more CASP3 apoptotic

activity as early as 5 post infection (p<0.01 paired t-test), 60% more total CASP3 activity, and greater

numbers of CASP3-positive cells in SOX2- (+70%), TBR2- (+65%), and CTIP2-positive cells (+60%) than in

mock-infected controls.

Following a high-content chemical screen, Zhou et al. (2017)31 discovered that Hiastrine

hydrobromide (HH) rescued a ZIKV-induced growth and differential defects in human forebrain-specific

organoids. They found an increase in viral copies after 48 hours, whereby ZIKV preferentially infected

more than 80% of SOX2-positive progenitor cells, which were intermixed with TUJ1-positive neuronal

cells and lacked discernible organization. There was a 10% decrease in Ki67-positive proliferating cells

(p<0.05 unpaired two-tailed t-test) and 20% increase in CASP3 apoptotic activity (p<0.001 unpaired two-

tailed), but no difference when ZIKV-infected organoids were treated with HH. Despite a 75% decrease

in area of ZIKV-infected organoid (p<0.001 unpaired two-tailed t-test) in comparison to mock-infected

controls, HH-treated organoids infected with ZIKV rescued ZIKV-induced growth deficits.

Sacramento et al. (2017)19 discovered that Sofosbuvir at 2 and 10 µM inhibited ZIKV replication

in human brain organoids (-91% and 99% respectively, p<0.01 ANOVA vs. 0 µM) and induced an

increased frequency of transitions (+37%) and A-to-G (+352%) ZIKV mutations (p<0.05 ANOVA vs. mock-

treated controls).

Li C et al. (2017)25, investigating the role of 25-Hydroxycholesterol (25HC) produced in the body,

discovered that it inhibited ZIKV infection in human cortical-specific organoids and has the potential as a

first-line antiviral agent to combat ZIKV. They demonstrated that the organoids permitted and increased

ZIKV replication as early as 12 hours (p<0.001 unpaired t-test vs. 0 hours). Concurrently, they found that

expression of the CH25H gene doubled after 12 hours following ZIKV infection (p<0.001 unpaired t-test

for all vs. 0 hours), and that 0.5 µM 25HC treatment decreased ZIKV replication by 75% (p<0.01 unpaired

t-test) in comparison to EtOH mock-treatment. There was also 18% decrease of ZIKV-positive cells in

25HC-treated cells (p<0.01 unpaired t-test vs. EtOH mock-treatment).

Synthesis of results

13

Viral replication and preferential infectivity. Six articles quantified ZIKV numbers from organoid

supernatant using qRT-PCR. There were four measurements of the African strain (MR766)22,23,27,31, three

of the Latin American strains (PRVABC5925, FB-GWUH-201623, and one isolated from a Brazilian

patient19), while only one from Asia (H/PF/201323). All showed increasing viral numbers after 48 and 72

hours. Furthermore, four articles studied the location of infection within the organoid, which showed

MR76623,27 of Africa, PRVABC5928 of Latin America, and FSS1302527 and H/PF/201323 from Asia

preferentially infected SOX2-, PAX6-, or phospho-vimentin-positive neural progenitor cells over CTIP2-

positive or TUJ1-positive neuronal cells, and followed lastly by TBR2-positive intermediate cells.

A comparison of strains by Qian et al. (2016)27 showed that the Asian strain was found

predominantly in SOX2-positive progenitor cells (90%) followed by CTIP2-positive neuronal cells (5%)

and TBR2-positive intermediate cells (5%), while the majority of the African strain was found in SOX2-

positive cells (92.5%) followed by CTIP2- (5%) and TBR2-positive cells (2.5%).

Cell population. Five articles showed that ZIKV of all three lineages reduced the neural

progenitor cell population20,21,22,23,28 in comparison to mock-infected controls, while three articles

revealed that ZIKV of all three lineages reduced the population of mature neurons21,23,28. Two articles

studying the African and Latin American strains reported a reduction in intermediate cell population21,28.

However, closer investigation revealed varying effects depending on strain. Cugola et al. (2016)21

showed greater reduction in SOX2-positive progenitor cell population after 96 hours for the Latin

American strain (-100%) than the African strain (-50%). Similarly, they also showed similar trends in

TBR1-positive intermediate cells (-75% for the Latin American strain, -50% for the African strain) and

CTIP2-positive neuronal cells (-75% for the Latin American strain, -25% for the African strain). On the

other hand, Gabriel et al. (2017)23 revealed greater reduction in phospho-vimentin-positive progenitor

cell population after 11 days of infection for the African strain (-93%) than either the Latin American or

Asian strain (-83% for both). Furthermore, the African strain was found to infect cells distant from

SOX2- or PAX6-positive cells and in greater numbers outside of the ventricular zone, whereas the Latin

American and Asian strains infected SOX2- and PAX6-positive cells located within the ventricular zone

and reduced TUJ1-positive cells at the organoid surface in similar fashion.

Zhou et al. (2017)31 observed that SOX2- and TUJ1-positive cells were intermixed and lacked

discernible organization in organoids infected by the African strain. Yoon et al. (2017)20 found abnormal

multipolar morphology in organoids expressing the non-structural protein of the African strain.

14

Cellular activity. Four articles investigated cellular activity by measuring the number of

proliferating cells. The African strain was shown to have decreased Ki67-20,21,31 and phosphate-H3-

positive27 proliferating cell populations in comparison to mock-infected controls, while the Latin

American strain recorded similar reductions in Ki67-positive cells21. There were no studies conducted

using strains from Asia. A comparison of lineages by Cugola et al. (2016)21 showed that the Latin

American strain had greater reduction after 96 hours over the African strain (-90% for the Latin

American strain, -80% for the African strain).

Gabriel et al. (2017)23 was the only article to investigate the effects of mitotic activity by ZIKV.

They found similar increases of dividing cells exhibiting vertical orientation of the division plane, an

indication of premature differentiation, after 5 and 11 days of ZIKV infection from all three lineages

(+150% after 5 days and +50% after 11 days for all lineages) when compared to mock-infected controls.

Cell death. Two studies measured cell death induced by the African strain using EdU activity.

Yoon et al. (2017)20 found that the co-expression of ZIKV-NS2A in forebrain organoids led to decreased

viability (-17.5%), while Qian et al. (2016)27 showed a decrease in viable cells within the ventricular zone

in infected organoids.

Seven articles measured apoptosis. All three lineages were shown to produce increased

CASP321,27,28,29,30 or TUNEL activity21,23 when compared to mock-infected controls. Comparisons of the

apoptotic effects according to lineage revealed varying results. While Cugola et al. (2016)21 showed

greater apoptotic activity by the Latin American strain than the African strain over 24 (+15% and +5%

respectively) and 96 hours post infection (+30% and +15% respectively) using TUNEL assay, CASP3 assay

revealed similar levels of apoptosis over the same time period (+500% for both at 24 hours, +1500% for

both at 96 hours). Gabriel et al. (2017)23 on the other hand, showed greater levels of apoptotic activity

using TUNEL assay for the African strain (+1100%) than the Latin American or Asian strains (+500%) after

11 days of infection. They also found increased TUNEL-positive cells in areas distinct from the

ventricular zone for African strain, as opposed to lineages from Latin America or Asia.

Wells et al. (2016)30 showed elevated CASP3 activity in response to increased viral presence

(R2=0.3599 for the Latin American strain), relative to mock-infected controls.

Organoid size. Nine articles reported smaller organoid size, area, volume, or thickness induced

by ZIKV of all lineages when compared to mock-infected controls21,22,23,24,26,27,28,29,31. Measurements were

recorded using either bright-field21,22,24,27,28,29,31, light sheet fluorescent26, or electron microscopy23. A

comparison of lineages by Gabriel et al. (2017)23 showed greater reduction in organoid diameter by the

15

African strain (-66%) than either the Latin American or Asian strains (-33% for both) in comparison to

mock-infected controls after 11 days of infection.

A comparison of articles using the African strain with the same duration of infection (24 hours)

and MOI (1) revealed that 10 day old ZIKV-infected organoids resulted in a 50% reduction in volume22,

whereas 6 week old ZIKV-infected organoids produced a 75% decrease in size26 when compared to

mock-infected controls. A shorter duration of ZIKV infection (2 hours) using the Latin American strain in

organoids of similar age (3 weeks) and exposed to similar MOI (1) showed a 25% reduction in area by

Watanabe et al. (2017)28 and a 35% decrease by Wells et al. (2016)29.

Qian et al. (2016)27 observed an enlarged lumen (+200%) with a reduction in ventricular zone

area (-75%) and thickness (-67%) by the African strain in comparison to mock-infected controls. Gabriel

et al. (2017)23 reported a greater reduction of ventricular zone thickness by the African strain (-100%)

than Latin American or Asian strains (-57% and -47% respectively). Furthermore, they measured a 43%

decrease in cortical plate thickness for the Asian strain, while Cugola et al. (2016)21 showed a 67% and

60% reduction in cortical plate thickness by the Latin American strain, as measured by TBR1- and CTIP2-

positive cells respectively. Wells et al. (2016)30 showed a decrease in area in response to increased MOI

for the Latin American strain.

Risk of bias across studies

There is currently no standardized protocol for ZIKV dilution or MOI during the infection process.

MOI ranged from values under 1 21,24,25,27,28,29,30,31 to as high as 1019,29 among all studies. Furthermore, at

present, there is no standardized duration of infection. Organoids were infected either for 2

hours19,24,25,28,29 or more than 24 hours21,22,23,26,27,30. Finally, as previously mentioned, ages for organoids

upon infection varied greatly, as there is currently no protocol as to the standardization of organoid

maturity. These variabilities may have influence the destruction caused by ZIKV on specific cells.

Discussion

Summary of evidence

As summarized in Table 4, it can be seen that all ZIKV strains, when grouped according to lineage

(i.e. African, Latin American, and Asian) had similar effects. In comparison to mock-infected controls,

16

the exposure of human brain organoids to different ZIKV strains resulted in the productive replication

after 72 hours, and preferentially infected neural progenitor cells over mature neurons. The

accumulation of ZIKV copies in organoids may indicate that neural progenitor cells served as viral

reservoir in the CNS32, while the low percentage of infected mature neurons suggested that neurons

may have inherited the virus from infected neural progenitor cells17. Therefore, deficits in neuronal and

intermediate progenitor cell populations may have been due to the depletion of neural progenitor cell

population during cortical neurogenesis. The heightened susceptibility of neural progenitor cells to ZIKV

infection was previously attributed to the expression of AXL, a member of the TAM/TIM family of

transmembrane receptors17,33. However, this was debunked by Wells et al. (2016)29 using genetically

modified AXL knockout human organoids, and further confirmed by mice studies34. It is still possible that

other members of the TAM/TIM family, such as TYRO3, MER, and TIM1, could contribute to ZIKV entry

into the cell as another organoid study by Watanabe et al. (2017)28 revealed that their expression

coincided with AXL in the neural progenitor cells within the ventricular and subventricular zones.

Further studies need to investigate these possible port of entry.

The review showed that all lineages of ZIKV reduced the cell populations of neural progenitor

cells and mature neurons, while limited data involving only African and Latin American lineages revealed

a reduction in the population of proliferating cells and intermediate cells. There was also premature

differentiation of neural progenitor cells and decreased overall viability. The potential mechanism by

which ZIKV reduce neural progenitor cell numbers in the developing cortex was proposed by the

organoid study from Yoon et al. (2017)20, in which ZIKV-NS2A interacted and degraded adherens junction

complex components, leading to deficits in neurogenesis and aberrant radial glial fiber scaffolding. On

the other hand, in another organoid study, Gabriel et al. (2017)23 remarked that neural progenitor cell

depletion may have been due to premature neural progenitor cell differentiation, caused by

centrosomal structural defects. Tang et al. (2016)35 revealed that ZIKV increased cell death and

dysregulated cell cycle progression, resulting in attenuated neural progenitor cell growth. Devhare et al.

(2017)36 reported that ZIKV dysregulated human neural stem cell growth and inhibited differentiation

into neural progenitor cells, whereas Souza et al. (2016)37 had shown that ZIKV induced mitosis

abnormalities and apoptotic cell death of human neural progenitor cells. Therefore, the cause of cell

cycle and neurogenesis dysregulation as induced by ZIKV must be investigated further.

The systematic review also revealed that all lineages of ZIKV caused heightened levels of

apoptosis. ZIKV has been shown to elicit P53 activation and apoptosis not only in cells expressing high

levels of viral antigens, but also in cells showing low or undetectable levels of the same protein38.

17

Currently unknown was whether the direct infection of neural progenitor cells was solely responsible for

increased cell death, as Gabriel et al. (2017)23 observed apoptosis occurring in areas distinct from the

ventricular zone, an area rich in neural progenitor cells and preferentially targeted by ZIKV, while Qian et

al. (2016)29 reported increased apoptosis in non-infected cells. A study by Bayless et al. (2016)39

observed that ZIKV-infected cranial neural crest cells promoted apoptosis of neural progenitor cell

population via a paracrine fashion. There were also differences in severity among the strains. Within

the comparative studies, Cugola et al. (2016)21 reported a greater percentage of apoptotic cells caused

by the Latin American strain than the African strain, while Gabriel et al. (2017)23 revealed the opposite:

that the African strain caused more apoptosis than the lineages from Asia or Latin America. Given that

the different strains are approximately 88.8% identical/97% amino acid40, further insight into the

molecular determinants of the disease should be investigated.

Finally, the systematic review revealed that all strains were responsible for a reduction in

organoid size. Thus, while the infection with different ZIKV strains promoted similar immature neural

progenitor cell death, the mechanisms by which they obtained this effect may likely be different. In

vitro studies comparing the mortality of neural progenitor cells following exposure to the African strain,

which is not associated with neural damage, or the Latin American and French Polynesian strains, both

associated with microcephaly, discovered that the latter induced stronger interferon response and P53

activation41. Innate immunity response through antiviral effectors such as type I and type II interferons

is thought to govern both viral replication and pathogenesis, yet aberrant activation of innate immunity

also results in inflammation, apoptosis, and autophagy, which might facilitate ZIKV spread and account

for the cytopathic effects in neural progenitor cells42,43. Using human organoids, Dang et al. (2016)22

found that ZIKV activated TLR3, an immune response receptor meditator, leading to perturbed cell fate,

and a reduction in organoid volume reminiscent of microcephaly. On the other hand, Chavali et al.

(2017)44 found that the ZIKV disrupted the binding of the neural RNA-binding protein Musashi-1, which is

highly expressed in neural progenitor cells, to its endogenous targets and dysregulated neural stem cell

function. Musashi-1 mutation is commonly found in individuals with autosomal recessive primary

microcephaly. Furthermore, a recent study found a Musashi-1 consensus binding site in the ZIKV

untranslated region of ZIKV lineages from Latin America and Asia, but not from Africa45. Therefore,

further investigations with organoids may prove rewarding.

This is the first systematic review of the effects of ZIKV on human brain organoids, which

support observations reported by studies using 2D cultures15,46, animal models47,48,49 and a clinical sample

of a ZIKV-infected fetus50. The emergence of a new experimental model during the present epidemic

18

finally allows for the opportunity to analyze if, in fact, ZIKV directly causes congenital microcephaly.

Qian et al., (2016)29 was the only study to show increased vulnerability in younger organoids, which

supports current observations that early first trimester infection causes greater neurological damage to

the fetus9. This is likely due to the fact that the ratio of neural progenitor cells to differentiated neurons

is greater during this period of development. The importance of determining if human brain organoids

can replicate the developing brain in utero is that it can be a powerful resource for the search of drugs

or vaccines to combat the ZIKV epidemic. For example, a potential ZIKV treatment, Azithromycin,

discovered with the use of 2D cultures51, showed limited protective effects with human brain

organoids30. There are several compounds that showed promising results in 2D culture experiments that

warrant further investigations using human brain organoids, such as a small-molecule

oligosaccharyltransferase inhibitor52, the repurposing of the anti-malaria drug chloroquine53, or the

development of an anti-ZIKV-NS1 antibody54. As the current organoid model improves, further

validation of its predictive value is merited.

Limitations and Future directions

All current reported studies evaluated the direct exposure of ZIKV to brain organoids in the

absence of immune and vascular systems. It has been shown that maternally infected ZIKV can cross the

mouse placental barrier, infect neural progenitor cells of the fetal brain, and cause a reduction in the

proliferative pool of cortical neural progenitors55. Thus, this might change the actual effects of ZIKV

infection in humans. However, future organoids may be able to provide natural viral delivery following

the development of protocols to incorporate hematopoietic lineages, endothelial cells56, and/or

microglia57.

As the emergence of organoids is relatively new, there is currently no gold standard protocol for

the generation of human brain organoids or MOI for viral infectivity. As well, the existence and use of

multiple ZIKV strains, despite having the same outcome of neural progenitor cell death and

microcephaly, have shown to use varying mechanisms that will undoubtedly limit the efficacy of ZIKV

research. One solution would be to require only the use of contemporary Latin American strains for

future research as those are the ones creating the most impact to society. Finally, in vitro experiments

involving human cerebral organoids are costly, thereby limiting its uptake as the new standard for ZIKV

research to only a handful of well-funded laboratories worldwide. However, as the use of organoids

expands, it is very likely that this new model will become more accessible in the future.

19

Conclusion

This systematic review strengthened the hypothesis that ZIKV causes congenital microcephaly,

as investigated using the human brain organoid model. In summary, sufficient evidence concluded that

ZIKV caused productive replication after 72 hours, preferentially infected neural progenitor cells over

mature neurons, reduced both cell populations, and caused premature differentiation. Limited data

involving only African and Latin American lineages also showed a reduction in populations of

proliferating cells and intermediate cells and overall decreased viability. Furthermore, all three strains

caused heightened apoptosis and the reduction in organoid size. Given the rapid pace of both ZIKV

research and organoid improvements, the future holds great promise for a therapeutic solution to this

terrible epidemic.

20

References

1. Yuan L, Huang XY, Liu ZY, Zhang F, Zhy XL, Yu JY, Ji X, Xu YP, Li G, Li C, Wang HJ, Deng YQ, Wu M, Chen ML, Ye Q, Xie DY, Li XF, Wang X, Shi W, Hu B, Shi PY, Xu Z, and Qin CF. (2017) A single mutation in the prM protein of Zika virus contributes to fetal microcephaly. Science. 358, 933-936.

2. Brasil P, Pereira JP, Moreira ME, Nogueira RMR, Damasceno L, Wakimoto M, Rabello RS, Valderramos SG, Halai UA, Salles TS, Zin AA, Horovitz D, Daltro P, Boechat M, Gabaglia CR, Carvalho de Sequeira P, Pilotto JH, Medialdea-Carrera R, Cotrim de Cunha D, Abreu de Carvalho LM, Pone M, Siqueira AM, Calvet GA, Baiao AER, Neves ES, Nassar de Carvalho P, Hasue RH, Marschik PB, Einspieler C, Janzen C, Cherry JD, Bispo de Filippis AM, and Nielsen-Saines K. (2016) Zika virus infection in pregnant women in Rio de Janeiro. The New England Journal of Medicine. 375, 2321-2334.

3. de Fatima Vasco Aragao M, van der Linden V, Brainer-Lima AM, Coeli RR, Rocha MA, Sobral da Silva P, Durce Costa Gomes de Carvalho M, van der Linden A, Cesario de Holanda A, and Valencia MM. (2016) Clinical features and neuroimaging (CT and MRI) findings in presumed Zika virus related congenital infection and microcephaly: retrospective case series study. BMJ. 353, i1901.

4. Anfasa F, Siegers JY, van der Kroeg M, Mumtaz N, Raj VS, de Vrij FMS, Widagdo W, Gabriel G, Salinas S, Simonin Y, Reusken C, Kushner SA, Koopmans MPG, Haagmans B, Martina BEE, and van Riel D. (2017) Phenotypic differences between Asian and African lineage Zika viruses in human neural progenitor cells. mSphere. 2(4), e00292-17.

5. Lazear HM, Govero J, Smith AM, Platt DJ, Fernandez F, Miner JJ, and Diamond MS. (2016) A mouse model of Zika virus pathogenesis. Cell Host Microbe. 19(5), 720-730.

6. Hansen DV, Lui JH, Parker PR, and Kriegstein AR. (2010) Neurogenic radial glia in the outer subventricular zone of human neocortex. Nature. 464, 554-451.

7. Lui JH, Hansens DV, and Kriegstein AR. (2011) Development and evolution of the human neocortex. Cell. 146, 18-36

8. Greschwind DH, and Rakic P. (2013) Cortical evolution : judge the brain by its cover. Neuron. 80, 633-647.

9. El Costa H, Gouilly J, Mansuy JM, Chen Q, Levy C, Cartron G, Veas F, Al-Daccak R, Izopet J, and Jabrane-Ferrat N. (2016) Zika virus reveals broad tissue and cell tropism during the first trimester of pregnancy. Scientific Reports. 6, i35296.

10. Salick MR, Wells MF, Eggan K, and Kaykas A. (2017) Modelling Zika virus infection of the developing human brain in vitro using stem cell derived cerebral organoids. Journal of Visualized Experiments. 127, e56404.

11. Qian X, Nguyen HN, Jacob F, Song H, and Ming G. (2017) Using brain organoids to understand Zika virus-induced microcephaly. Development. 144, 952-957

12. Bae BI, Jayaraman D, and Walsh CA. (2015) Genetic changes shaping the human brain. Developmental Cell. 32, 423-434.

13. Diaz AL, and Gleeson JG. (2009) The molecular and genetic mechanisms of neocortex development. Clinical Perinatology. 36, 503-512.

14. Sun T and Hevner RF. (2014) Growth and folding of the mammalian cerebral cortex: from molecules to malformations. Nature Review Neuroscience. 15, 217-232.

15. Onorati M, Li Z, Liu F, Sousa AMM, Nakagawa N, Li M, Dell'Anno MT, Gulden FO, Pochareddy S, Tebbenkamp ATN, Han W, Pletikos M, Gao T, Zhu Y, Bichsel C, Varela L, Szigeti-Buck K, Lisgo S, Zhang Y, Testen A, Gao XB, Mlakar J, Popovic M, Flamand M, Strittmatter SM, Kaczmarek LK, Anton ES, Horvath TL, Lindenbach BD, and Sestan N. (2016) Zika virus disrupts phospho-TBK1

21

localization and mitosis in human neuroepithelial stem cells and radial glia. Cell Reports. 16, 2576-2592.

16. Lancaster MA, Renner M, Martin CA, Wenzel D, Bicknell LS, Hurles ME, Homfray T, Penninger JM, Jackson AP, and Knoblich JA. (2013) Cerebral organoids model human brain development and microcephaly. Nature. 501, 373-379.

17. Nowakowski TJ, Pollen AA, Di Lullo E, Sandoval-Espinosa C, Bershteyn M, and Kriegstein AR. (2016) Expression analysis highlights AXL as a candidate Zika virus entry receptor in neural stem cells. Cell Stem Cell. 18, 591-596.

18. Zhu Z, Gorman MJ, McKenzie LD, Chai JN, Hubert CG, Prager BC, Fernandez E, Richner JM, Zhang R, Shan C, Tucksen R, Wang X, Shi PY, Diamond MS, Rich JN, and Chheda MG. (2017) Zika virus has oncolytic activity against glioblastoma stem cells. The Journal of Experimental Medicine. 214(10), 2843-2857.

19. Sacramento CQ, de Melo GR, de Freitas CS, Rocha N, Hoelz LVB, Miranda M, Fintelman-Rodrigues N, Marttorelli A, Ferreira AC, Barbosa-Lima G, Abrantes JL, Vieira TR, Bastos MM, Volotao EM, Nunes EP, Tschoeke DA, Leomil L, Loiola EC, Trindade P, Rehen SK, Bozza FA, Bozza PT, Boechat N, Thompson FL, de Filippis AMB, Bruning K, and Souza TML. (2017) The clinically approved antiviral drug sofosbuvir inhibits Zika virus replication. Scientific Reports. Srep40920.

20. Yoon KJ, Song G, Qian X, Pan J, Xu D, Rho HS, Kim NS, Habela C, Zheng L, Jacob F, Zhang F, Lee EM, Huang WK, Ringeling FR, Vissers C, Li C, Yuan L, Kang K, Kim S, Yeo J, Cheng Y, Liu S, Wen Z, Qin CF, Wu Q, Christian KM, Tang H, Jin P, Xu Z, Qian J, Zhu H, Song H, and Ming GL. (2017) Zika-virus-encoded NS2A disrupts mammalian cortical neurogenesis by degrading adherens junction proteins. Cell Stem Cell. 21, 349-358.

21. Cugola FR, Fernandes IR, Russo FB, Freitas BC, Dias JLM, Guimaraes KP, Benezzato C, Almeida N, Pignatari GC, Romero S, Polonio CM, Cunha I, Freitas CL, Brandao WN, Rossato C, Andrade DG, Faria DP, Garcez AT, Buchpigel CA, Braconi CT, Mendes E, Sall AA, Zanotto PMA, Peron JPS, Muotri AR, and Beltrao-Braga PCB. (2016) The Brazilian Zika virus strain causes birth defects in experimental models. Nature. 534, 267-271.

22. Dang J, Tiwari SK, Lichinichi G, Qin Y, Patil VS, and Eroshkin AM. (2016) Zika virus depletes neural progenitors in human cerebral organoids through activation of the innate immune receptor TLR3. Cell Stem Cell. 19, 258-265.

23. Gabriel E, Ramani A, Karow U, Gottardo M, Natarajan K, Gooi LM, Goranci-Huzhala G, Krut O, Peters F, Nikolic M, Kuivanen S, Korhonen E, Smura T, Vapalahti O, Papantonis A, Schmidt-Chanasit J, Riparbelli M, Callaini G, Kronke M, Utermohlen O, and Gopalakrishnan J. (2017) Recent virus isolates induce premature differentiation of neural progenitors in human brain organoids. Cell Stem Cell. 20, 397-406.

24. Garcez PP, Loiola EC, da Costa RM, Higa LM, Trindade P, Delvecchio R, Nascimento JM, Brindeiro R, Tanuri M, and Rehen SK. (2016) Zika virus impairs growth in human neurospheres and brain organoids. Science. 352(6287), 816-818.

25. Li C, Deng YQ, Wang S, Ma F, Aliyari R, Huang XY, Zhang NN, Watanabe M, Dong HL, Liu P, Li XF, Ye Q, Tian M, Hong S, Fan J, Zhao H, Li L, Vishlaghi N, Buth JE, Au C, Liu Y, Lu N, Du P, Qin FXF, Zhang B, Gong D, Dai X, Sun R, Novitch BG, Xu Z, Qin CF, and Cheng G. (2017) 25-Hydroxycholesterol protects host against Zika virus infection and its associated microcephaly in a mouse model. Immunity. 46, 446-456.

26. Li Y, Muffat J, Omer A, Bosch I, Lancaster MA, Sur M, Gehrke L, Knoblich JA, an Jaenisch R. (2017) Induction of expansion and folding in human cerebral organoids. Cell Stem Cell. 20, 385-396.

27. Qian X, Nguyen HN, Song MM, Hadiono C, Ogden SC, Hammack C, Yao B, Hamersky G, Jacob F, Zhong C, Yoon K, Jeang W, Lin L, Li Yujing, Thakor J, Berg D, Zhang C, Kang E, Chickering M,

22

Naeun D, Ho CY, Wen Z, Christian KM, Shi PY, Maher BJ, Wu H, Jin P, Tang H, Song H, and Ming G. (2016) Brain-region-specific organoids using mini-bioractors for modeling ZIKV exposure. Cell. 165, 1238-1254.

28. Watanabe M, Buth JE, Vishlaghi N, de la Torre-Ubieta L, Taxidis J, Khakh BS, Coppola G, Pearson CA, Yamauchi K, Gong D, Dai X, Damoiseaux R, Aliyari R, Liebscher S, Schenke-Layland K, Caneda C, Huang EJ, Zhang Y, Cheng G, Geschwind DH, Golshani P, Sun R, and Novitch BG. (2017) Self-organized cerebral organoids with human-specific features predict effective drugs to combat Zika virus infection. Cell Reports. 21, 517-532.

29. Wells MF, Salick MR, Wiskow O, Ho DJ, Worringer KA, Ihry RJ, Kommineni S, Bilican B, Klim JR, Hill EJ, Kane LT, Ye C, Kaykas A, and Eggan K. (2016) Genetic ablation of AXL does not protect human neural progenitor cells and cerebral organoids from Zika virus infection. Cell Stem Cell. 19, 703-708.

30. Xu M, Lee EM, Wen Z, Cheng Y, Huang Y, Qian X, Julia TCW, Kouznetsova J, Ogden SC, Hammack C, Jacob F, Nguyen HN, Itkin M, Hanna C, Shinn P, Allen C, Michael SG, Simeonov A, Huang W, Christian KM, Goate A, Brennand KJ, Huang R, Xia M, Ming GL, Zheng W, Song H, and Tang H. (2016) Identification of small-moleculr inhibitors of Zika virus infection and induced neural cell death via drug repurposing screen. Nature Medicine. 22(10), 1101-1107.

31. Zhou T, Tan L, Cederquist GY, Fan Y, Hartley BJ, Mukherjee S, Tomishima M, Brennand KJ, Zhang Q, Schwartz RE, Evans T, Studer L, and Chen S. (2017) High-content screening in hPSC-neural progenitors identifies drug candidates that inhibit Zika virus infection in fetal-like organoids and adult brain. Cell Stem Cell. 21, 274-283.

32. Nguyen HN, Qian X, Song H, and Ming GL. (2016) Neural stem cells attacked by Zika virus. Cell Research. 26, 753-754.

33. Retallack H, Di Lullo E, Arias C, Knopp KA, Laurie MT, Sandoval-Espinosa C, Leon WRM, Krencik R, Ullian EM, Spatazza J, Pollen AA, Mandel-Brehm C, Nowakowski TJ, Kriegstein AR, and LeRisi JL. (2016) Zika virus cell tropism in the developing human brain and inhibition by azithromycin. PNAS. 113(50, 14408-14413.

34. Hastings AK, Yockey LJ, Jagger BW, Hwang J, Uraki R, Gaitsch HF, Parnell LA, Cao B, Mysorekar I, Rothlin CV, Erol F, Diamond DS, and Iwasaki A. (2017) TAM receptors are not required for Zika virus infection in mice. Cell Reports. 19(3), 558-568.

35. Tang H, Hammack C, Ogden SC, Wen Z, Qian X, Li Y, Yao B, Shin J, Zhang F, Lee EM, Christian KM, Didier RA, Jin P, Song H, and Ming G. (2016) Zika virus infects human cortical neural precursors and attenuates their growth. Cell Stem Cell. 18(5), 587-590.

36. Devhare P, Meyer K, Steele R, Ray RB, and Ray R. (2017) Zika virus infection dysregulates human neural stem cell growth and inhibits differentiation into progenitor cells. Cell Death and Disease. 8, e3106.

37. Souza BSF, Sampaio GLA, Pereira CS, Campos GS, Sardi SI, Freitas LAR, Figueira CP, Paredes BD, Nonaka CKV, Azevedo CM, Rocha VPC, Bandeira AC, Mendez-Otero R, dos Santos RR, and Soares MBP. (2016) Zika virus infection induces mitosis abnormalities and apoptotic cell death of human neural progenitor cells. Scientific Reports. 6, 39775.

23

38. El Ghouzzi V, Bianchi FT, Molineris I, Mounce BC, Berto GE, Rak M, Lebon S, Aubry L, Tocco C, Gai M, Chiotto AMA, Sgro F, Pallavicini G, Simon-Loriere E, Passemard S, Vignuzzi M, Gressens P, and Di Cunti F. (2016) Zika virus elicits p53 activation and genotoxic stress in human neural progenitors similar to mutations involved in severe forms of genetic microcephaly and p53. Cell Death and Disease. 7, e2567.

39. Bayless NL, Greenberg RS, Swigut T, Wysocka J, and Blish CA. (2016) Zika virus infection induces cranial neural crest cells to produce cytokines at levels detrimental for neurogenesis. Cell Host & Microbe. 20, 423-428.

40. Dowall SD, Graham VA, Rayner E, Hunter L, Atkinson B, Pearson G, Dennis M, and Hewson R. (2017) Lineage-dependent differences in the disease progression of Zika infection in type-I interferon receptor knockout (A129) mice. PLOS Neglected Tropical Diseases. 11(7), e0005704.

41. Zhang F, Hammack C, Ogden SC, Cheng Y, Lee EM, Wen Z, Qian X, Nguyen HN, Li Y, Yao B, Xu M, X T, Chen L, Wang Z, Feng H, Huang WK, Yoon KJ, Shan C, Huang L, Qin Z, Christian KM, Shi PY, Xu M, Xia M, Zheng W, Wu H, Song H, Tang H, Ming GL, and Jin P. (2016) Molecular signatures associated with ZIKV exposure in human cortical neural progenitors. Nucleic Acids Research. 44(18), 8610-8620.

42. Liang Q, Luo Z, Zeng J, Chen W, Foo SS, Lee SA, Ge J, Wang S, Goldman SA, Zlokovic BV, Zhao Z, and Jung JU. (2016) Zika virus NS4A and NS4B proteins deregulate Akt-mTOR signaling in human fetal neural stem cells to inhibit neurogenesis and induce autophagy. Cell Stem Cell. 19, 663-671.

43. Chaudhary V, Yuen KS, Chan JFW, Chan CP, Wang PH, Cai JP, Zhang S, Liang M, Kok KH, Chan CP, Yuen KY, and Jin DY. (2017) Selective activation of type II interferon signaling by Zika virus NS5 protein. Journal of Virology. 91(14), E00163-17.

44. Chavali PL, Stojic L, Meredith LW, Joseph N, Nahorski MS, Sanford TJ, Sweeney TR, Krishna BA, Hosmillo M, Firth AE, Bayliss R, Marcelis CL, Lindsay S, Goodfellow I, Woods CG, and Gergely F. (2017) Neurodevelopmental protein Musashi-1 interacts with the Zika genome and promotes viral replication. Science. 357, 83-88.

45. Klase ZA, Khakhina S, Schneider ADB, Callahan MV, Glasspool-Malone J, and Malone R. (2016) Zika fetal neuropathogenesis: Etiology of a viral syndrome. PloS Neglected Tropical Diseases. 10(8), E0004877.

46. Meertens L, Labeau A, Dejarnac O, Cipriani S, Sinigaglia L, Bonnet-Mladin L, Charpentier TF, Hafirassou ML, Zamborlini A, Cao-Lormeau V-M, Coulpier M, Misse D, Jouvenet N, Tabibiazar R, Gressens P, Schwartz O, and Amara A. (2017) Axl mediates Zika virus entry in human glial cells and modulates innate immune responses. Cell Reports. 18, 324-333.

47. Yockey LJ, Varela L, Rakib T, Khoury-Hanold W, Finks SL, Stutz B, Szigeti-Buck K, van den Pol A, Lindenbach BD, Horvath TL, and Iwasaki A. (2016) Vaginal exposure to Zika virus during pregnancy leads to fetal brain infection. Cell. 166, 1247-1256.

48. Mlakar J, Korva M, Tul N, Popovic M, Poljsak-Prijatelj M, Mraz J, Kolenc M, Resman Rus K, Vesnavar Vipotnik T, Fabjan Vodusek V, Vizjak A, Pizem J, PEtrovec M, and Avsic Zupanc T. (2016) Zika virus associated with microcephaly. The New England of Medicine. 374, 951-958.

49. Shao Q, Herrlinger S, Yang SL, Lai F, Moore JM, Brindley MA, and Chen JF. (2016) Zika virus infection disrupts neurovascular development and results in postnatal microcephaly with brain damage. Development143, 4127-4136.

50. Ho CY, Ames HM, Tipton A, Vezina G, Liu JS, Scafidi J, Torii M, Rodriguez FJ, du Plessis A, and DeBiasi RL. (2017) Differential neuronal susceptibility and apoptosis in congenital ZIKV infection. Annals of Neurology. 82(1), 121-127.

51. Rowe RG, Mandelbaum J, Zon LI, and Daley GQ. (2016) Engineering hematopoietic stem cells : lessons from development. Cell Stem Cell. 18, 707-720.

24

52. Puschnik AS, Marceau CD, Ooi YS, Majzoub K, Rinis N, Contessa JN, and Carette JE. (2017) A small-molecule oligosaccharyltransferase inhibitor with pan-flaviviral activity. Cell Reports. 21, 3032-3039.

53. Shiryaev SA, Mesci P, Pinto A, Fernandes I, Sheets N, Shresta S, Farhy C, Huang CT, Strongin AY, Muotri AR, and Tersikh AV. (2017) Repurposing of the anti-malaria drug chloroquine for Zika virus treatment and prophylaxis. Scientific Reports. 7(1), 15771.

54. Wang D, Chen C, Liu S, Zhou H, Yang K, Zhao Q, Ji X, Chen C, Xie W, Wang Z, Mi L, and Yang H. (2017) A mutation identified in neonatal microcephaly destabilizes Zika virus NS1 assembly in vitro. Scientific Reports. 7, 42580.

55. Wu KY, Zuo GL, Li XF, Ye Q, Deng YQ, Huang XY, Cao WC, Qin CF, and Luo ZG. (2016) Vertical transmission of Zika virus targeting the radial glial cells affects cortex development of offspring mice. Cell Research. 26, 645-654.

56. Lippmann ES, Azarin SM, Kay JE, Nessler RA, Wilson HK, Al-Ahmad A, Palecek SP, and Shusta EV. (2012) Derivation of blood-brain barrier endothelial cells from human pluripotent stem cells. Nature Biotechnology. 30, 783-791.

57. Muffatt J, Li Y, Yuan B, Mitalipova M, Omer A, Corcoran S, Bakiasi G, Tsai LH, Aubourg P, Ransohoff RM, and Jaenisch R. (2016) Efficient derivation of microglia-like cells from human pluripotent stem cells. Nature Medicine. 22, 1358-1367.

25

Fig 1. Flow diagram of literature search and selection criteria adapted from PRISMA protocol.

26

Table 1. Summary of descriptive characteristics of included articles (n=13).

AuthorYear of Publication Country Study Design

Experimental model

Age of organoid ZIKV strain

Multiplicity of Infection (MOI) or dilution

Duration of infection Main conclusion

Cugola FR 2016 Brazil In vitro experimentation

Human cerebral organoids

28 days MR766 (ZIKV-AF)Brazilian (ZIKV-BR)

MOI 0.1 24 hours The infection of human brain organoids by ZIKV-BR or ZIKV-AF results in a reduction of proliferative zones and disrupted cortical layers. The data reinforces evidence linking ZIKV-BR outbreak to congenital brain malformations.

Dang J 2016 USA In vitro experimentation

Human cerebral organoids

10 days MR766 (ZIKV-AF)

MOI 1 24 hours ZIKV strain MR766 efficiently infects organoids and causes a decrease in overall organoid size that correlates with the kinetics of viral copy number. A linked is identified between ZIKV-mediated TLR3 activation, perturbed cell fate, and a reduction in organoid volume reminiscent of microcephaly.

Gabriel E 2017 Germany In vitro experimentation

Human brain organoids

9 days MR766FB-GWUH-2016 (ZIKV-AM)H/PF/2013 (ZIKV-AS)

MOI 1 2 days, 5 days, or 11 day

Two recently isolated strains ZIKV-AM and ZIKV-AS efficiently infect neural progenitors in human brain organoids, causing premature differentiation in a way closely resembling ZIKV-associated microcephaly. The effects are different than that seen with ZIKV-AF, an older and extensively passaged strain.

27

Garcez P 2016 Brazil In vitro experimentation

Human brain organoids

35 days MR766 MOI 0.0025-0.25 (1:100, 1:1,000 dilution)

2 hours ZIKV targets human brain cells, reducing their viability and growth as brain organoids, suggesting that ZIKV abrogates neurogenesis during human brain development.

Li C 2017 China In vitro experimentation

Human cortical-specific organoids

23 days PRVABC59 MOI 0.3-3.12 (1:2-1:8 dilution)

2 hours 25HC (25-Hydroxycholesterol inhibits ZIKV infection in human cortical organoids and has the potential as a first-line antiviral agent to combat ZIKV.

Li Y 2017 USA In vitro experimentation

Human cerebral organoids

19 days or 30 days

MR766 MOI 1 24 hours Infection of cerebral organoids with ZIKV impairs cortical growth and folding.

Qian X 2016 USA In vitro experimentation

Human forebrain-specific organoids

14 days or 28 days

MR766FSS13025

1:10 dilution for 14 day organoids1:10 or 1:40 dilution for 28 day organoids

24 hours Using a brain-region-specific organoid, there is preferential, productive infection of neural progenitors with either African or Asian ZIKV strains leading to increased cell death and reduced proliferation, resulting in decreased neuronal cell-layer volume resembling microcephaly.

Sacramento CQ

2017 Brazil In vitro experimentation

Human brain organoids

Not available

Brazilian MOI 1.0 or 10 2 hours Sofosbuvir inhibited ZIKV replication in brain organoids and induced an increase in A-to-G mutations in the viral genome.

28

Watanabe M 2017 USA In vitro experimentation

Human cortical-specific organoids

21 days or 56 days

PRVABC59 MOI 0.3-3.125 (1:2-1:8 dilution)

2 hours An optimized organoid culture method that efficiently and reliably produce cortical and basal ganglia structures similar to those in the human fetal brain in vivo models the teratogenic effects of ZIKV on the developing brain.

Wells MF 2016 USA In vitro experimentation

Human cerebral organoids

24 days PRVABC59 MOI 0.1, 1, or 10

2 hours Genetic ablation of AXL has no effect on ZIKV entry or ZIKV-mediated cell death in cerebral organoids.

Xu M 2016 USA In vitro experimentation

Human forebrain-specific organoids

18 days FSS13025 MOI 0.04 to 0.08

24 hours A pan-caspase inhibitor, Emricasan, inhibited ZIKV-induced increases in caspase-3 activity and protected three-dimensional organoid cultures.

Yoon KJ 2017 USA In vitro experimentation

Human forebrain-specific organoids

45 days MR766 Not available Co-expression of ZIKV NS2A in radial glia cells

Systematically introducing individual proteins encoding ZIKV-NS2 in human forebrain organoids reduces radial glial cell proliferation and causes adherens junction deficits.

Zhou T 2017 USA In vitro experimentation

Human forebrain-specific organoids

20 days MR766 MOI 0.125 24 hours Hiastrine hydrobromide rescued a ZIKV-induced growth and differential defects in human fetal-like forebrain organoids.

29

Table 2. Summary of assays and results for included articles (n=13).

Author Methods: Assays ResultsCugola FR Marginal zone, cortical

plate, and ventricular zone staining for CTIP2, TBR1, MAP2, TUJ1, Ki67, PAX6, SOX2, CASP3, or TUNEL in ZIKV-BR-infected, ZIKV-AF-infected, YFV-infected, or mock-infected human cerebral organoids. Percentage of TBR1-positive cells, CTIP2-positive cells, PAX6-positive cells, Ki67-positive cells, SOX2-positive cells, CASP3-positive cells, or TUNEL-positive cells in ZIKV-BR, ZIKV-AF, or YFV-infected organoids vs. mock-infected controls at 24 and 96 hours. Measurement of cortical thickness by extension of TBR-1 positive or CTIP2-positive cell layer. Nuclear size of CASP3-positive cells.

After 96 hours infection, decreased cortical plate thickness of TBR1-positive cells (-67%, p<0.05 ANOVA vs. mock-infected controls) and CTIP2-positive cells (-60%, p<0.05 ANOVA vs. mock-infected controls.) for ZIKV-BR. Decreased SOX2-positive cells after 24 hours and 96 hours infection for both ZIKV-BR (-37.5% and -100% respectively) and ZIKV-AF (-25% and -50% respectively). Decreased PAX6-positive cells after 96 hours infection for ZIKV-BR (-50%). Decreased PAX6-positive cells after 24 hours and 96 hours infection for ZIKV-AF (-25% and -75%). Decreased Ki67-positive cells after 24 hours and 96 hours infection for both ZIKV-BR (-25% and -90% respectively) and ZIKV-AF (-60% and -80% respectively). Decreased TBR1-positive cells after 24 hours and 96 hours infection for both ZIKV-BR (-15% and -75% respectively) & ZIKV-AF (-25% and -50% respectively). Increased CTIP2-positive cells after 24 hours infection for ZIKV-BR (+25%) and ZIKV-AF (+12.5%), but decreased CTIP2-positive cells after 96 hours infection for ZIKV-BR (-75%) and ZIKV-AF (-25%). Increased TUNEL-positive cells after 24 hours and 96 hours infection by ZIKV-BR (+15% and +30% respectively) and ZIKV-AF (+5% and 15% respectively). Increased CASP3 activity after 24 hours and 96 hours infection for both ZIKV-BR (+500% and +1500% respectively) and ZIKV-AF (+500% and +750% respectively).

Dang J Bright-field images with quantification of size of ZIKV-infected or mock-infected individual human cerebral organoids over time. ZIKV viral copy count in organoid supernatant. Staining for ZIKV envelope protein and Nestin. RT-qPCR analysis of TLR3 expression, genes Ntn1 and Ephb2, in mock-infected and ZIKV-infected organoids, with and without a TLR3 competitive inhibitor.

After 5 days infection, decreased volume (-50%, p<0.05 t-test vs. mock-infected controls). Decreased NESTIN-positive cells after ZIKV infection. Increased ZIKV viral copies after 24 and 48 hours (+10,000% and +316,200% respectively). Upregulation of TLR3 mRNA levels in ZIKV-infected organoids (+75%). Downregulation of Ntn1 (-100%, p<0.01 vs. mock-infected controls) and Ephb2 (-80%, p<0.05 t-test vs. mock-infected controls) genes in ZIKV-infected organoids. Upregulation of Ntn1 (+250%) and Ephb2 (+1000%, p<0.001 t-test vs. ZIKV-infected organoids) genes in ZIKV-infected organoids with TLR3 competitive inhibitor.

Gabriel E ZIKV-AM, ZIKV-AS, and ZIKV-AM viral count in organoid supernatant. Electron microscopy images with quantification of size of ZIKV-AM, ZIKV-AS, ZIKV-AF, or mock-infected organoids over time. Cortical plate and ventricular zone staining for TUJ1, SOX2, Pax6, Nestin, DCX, MAP2, AXL, Tunel, and phospho-vimentin/Arl13b (division plane orientation).

After 5 days infection, decreased cortical plate thickness for ZIKV-AS (-43%, p<0.001 one way ANOVA vs. mock-infected controls), decreased ventricular zone thickness for ZIKV-AS (-13%, p<0.01 one way ANOVA vs. mock-infected controls), and ZIKV-AM (-26%, p<0.001 one way ANOVA vs. mock-infected controls). After 11 days infection, decreased diameter for ZIKV-AF (-66%, p<0.001 one way ANOVA vs. mock-infected controls), ZIKV-AS (-33%, p<0.001 one way ANOVA vs. mock-infected controls, and ZIKV-AM (-33%, p<0.001 one way ANOVA vs. mock-infected controls), decreased ventricular zone thickness for ZIKV-AF (-100%, p<0.05 one way ANOVA vs. mock-infected controls), ZIKV-AS (-47%, p<0.001 one way ANOVA vs. mock-infected controls), and ZIKV-AM (-57%, p<0.001 one way ANOVA vs. mock-infected controls). Greater number of ZIKV-AS and ZIKV-AM-positive cells in the ventricular zone (NPCs) than outside the ventricular zone. Greater number of ZIKV-AF cells outside the ventricular zone than inside the ventricular zone. Increased number of ZIKV-AS (+50%) and ZIKV-AM-infected cells (+43%) at 11 days post infection than vs. 5 days post infection, whereas decreased number of ZIKV-AF-positive cells at 11 days post infection vs. 5 days post infection (-100%). Decreased phoso-vimentin-positive cells for ZIKV-AS, ZIKV-AM, and ZIKV-AF at 5 days (-66% for all three , p<0.001 vs. mock-infected controls) and 11 days (-83%, -83%, and -93% respectively, p<0.001 vs. mock-infected controls) post infection. Decreased percentage of cells showing horizontal orientation for ZIKV-AF, ZIKV-AS, and ZIKV-AM at 2 days (-62.5%, -50%, and -62.5%

30

Organoid diameter, cortical plate thickness, ventricular zone thickness.

respectively, p<0.001 two way ANOVA vs. mock-infected controls), 5 days (-60%, 67%, and 60% respectively, p<0.001 two way ANOVA vs. mock-infected controls), and 11 days (-42%, -50%, and -50% respectively, p<0.001 two way ANOVA vs. mock-infected controls). Increased percentage of cells showing vertical orientation for ZIKV-AF, ZIKV-AS, and ZIKV-AM at 2 days (+125%, +175%, and +200% respectively, p<0.001 two way ANOVA vs. mock-infected controls), 5 days (+150% for all three, p<0.001 two way ANOVA vs. mock-infected controls), and 11 days (+50% for all three, p<0.001 two way ANOVA vs. mock-infected controls) post infection. Increased apoptosis for ZIKV-AF, ZIKV-AS, and ZIKV-AM at 2 days (+300%, +60%, and +100% respectively), 5 days (+900%, +100%, and +220% respectively), and 11 days (+1100%, 500%, and 500% respectively) post infection. Increased percentage of apoptotic cells for ZIKV-AF, ZIKV-AS, and ZIKV-AM (+35%, +35%, and +25% respectively, p<0.001 one way ANOVA vs. mock-infected controls). Increased percentage of cells infected by ZIKV-AF (30%), ZIKV-AS (35%), and ZIKV-AM (40%). SOX2-positive cells infected by ZIKV-AS & ZIKV-AM (ZIKV-AF infected cells are distant from SOX2-positive cells). PAX6-positive cells in ventricular zone infected by ZIKV-AS & ZIKV-AM (ZIKV-AF infected cells are distant from PAX6-positive cells). Decreased TUJ1-positive cells at surface for ZIKV-AS and ZIKV-AM. Increased percentage of ventricular zone lumen exhibiting TUJ1-positive differentiated cells for ZIKV-AS (+900%, p<0.001 one way ANOVA vs. mock-infected controls) & ZIKV-AM (+1000%, p<0.001 one way ANOVA vs. mock-infected controls.) Increased TUNEL-positive cells in areas distinct from ventricular zone for ZIKV-AF (p<0.001 one way ANOVA vs. mock-infected controls).

Garcez P Bright-field images with quantification of size of ZIKV, DENV2, or mock-infected organoids over time. Organoid area.

Decreased area (-40%, p<0.05 unpaired two-tailed t-test vs. mock-infected controls).

Li C ZIKV viral copy count in organoid supernatant. qRT-PCR quantification of IFN-B and CH25H.

Increased ZIKV replication after 12, 48, and 72 hours (p<0.001 unpaired t-test vs. 0 hours). Increased expression of IFN-B and CH25H genes after 12 (+100% for both), 48 (+225% for both), and 72 hours (+200% for both) post ZIKV infection (p<0.001 unpaired t-test for all vs. 0 hours). Decreased ZIKV replication after 25HC treatment for 0.1 (-40%, p<0.01 unpaired t-test), 0.5 (-75%, p<0.01 unpaired t-test), and 2.5 (-99%, p<0.001 unpaired t-test) uM vs. EtOH mock-treatment. Decreased percentage of ZIKV-positive cells in 25HC-treated cells (-18%, p<0.01 unpaired t-test vs. EtOH mock-treatment).

Li Y Light sheet fluorescent microscopy and quantification of size and fold density of ZIKV and mock-infected organoids, both control and PTEN mutant. Staining for CASP3 and phospho-H3

Decreased size in control and PTEN mutant organoids by ZIKV (-75% and -82% respectively, p<0.001 t-test vs. mock-infected controls). Decreased fold density in PTEN organoids only by ZIKV (-94%, p<0.001 t-test vs. mock-infected controls). Increased apoptosis for 19 day & 30 day organoids.

Qian X Bright field images and quantification of organoid area, area of ventricular zone and lumen, and thickness of ventricular zone. Staining for CASP3, phospho-H3, SOX2, CTIP2, TUJ1, EdU, and TBR2.

Decreased organoid area (-83%, p<0.0005 t-test vs. mock-infected controls), area of ventricular zone (-75%, p<0.005 t-test vs. mock-infected controls), and thickness of ventricular zone (-67%, p<0.0005 vs. mock-infected controls) in ZIKV-M-infected organoids. Increased area of lumen (+200, p<0.05 vs. mock-infected controls) in organoids infected at day 14 and analyzed 18 days later. Decreased organoid area (-67%, p<0.0005 vs. mock-infected controls) in organoids infected at day 28 by ZIKV-M and analyzed 14 days later. Decreased thickness of neuronal layer (-50% for ZIKV-M 0.25x, p<0.005 t-test vs. mock-infected controls; -67% for ZIKV-M 1x, p<0.0005 t-test vs. mock-infected controls) and ventricular zone (-33% for ZIKV-M 0.25x, p<0.05 t-test vs. mock-infected controls; -50% for ZiKV-M 1x, p<0.0005 t-test vs. mock-infected controls) in ZIKV-M-infected organoids. Increased percentage of ZIKV-M envelope-positive cells from day 32 to day 42 for ZIKV-M-infected organoids (+40% for ZIKV 0.25x, +50% for ZIKV 1x). Majority of ZIKV-C enveloped-positive cells found in SOX2-positive cells (90%) followed by CTIP2 (5%) and TBR2-positive cells (5%). Majority of ZIKV-M enveloped-positive cells found in SOX2-positive cells (92.5%) followed by CTIP2 (5%) and TBR2-positive cells (2.5%). 35% of SOX2-positive cells and 10% of CTIP2-positive cells infected by ZIKV-M envelope. Increased percentage of CASP3-positive cells (+55%, p<0.0005 t-test vs. mock-infected controls) and decreased density of phospho-H3-positive cells (-87.5%, p<0.0005 t-test vs. mock-infected controls) in ZIKV-M-infected organoids. Decreased percentage of EdU-positive cells in ventricular zone (-28% for ZIKV-M 0.25x, -88% for ZIKV-M 1x, p<0.005 t-test vs. mock-infected controls for both). Increased CASP3-positive cells (+800% for ZIKV-M 0.25x, +17,000% for ZIKV-M 1x, p<0.0005 t-test vs. mock-infected controls for both).

Sacramento CQ ZIKV viral copy count in organoid supernatant. Observational analysis of mutations in ZIKV genome associated with Sofosbuvir treatment.

Sofosbuvir (2 and 10 uM) inhibits ZIKV replication (-91% and 99% respectively, p<0.01 ANOVA vs. 0 uM). Increased frequency of transitions (+37%) and A-to-G (+352%) ZIKV mutations (p<0.05 ANOVA vs. mock-treated controls).

Watanabe M Staining for SOX2, TBR2, CTIP2, PI/HOECHST, and CASP3. Bright field images and quantification

Decreased area (-14%, p<0.001 paired t-test vs. mock-infected controls) and perimeter (-20%, p<0.01 paired t-test vs. mock-infected controls) of ZIKV-infected organoids. ZIKV predominantly infects SOX2 (70%, of which 70% are infected), followed by TBR2 (20%, of which 70% are infected), and CTIP2-positive cells (10%, of which 15% are infected). Percentage of ZIKV-infection in SOX2 (95%), TBR2 (80%), and CTIP2-positive cells (60%) with ZIKV

31

of area and perimeter. envelope staining. Increased cell death in ZIKV-infected organoids (p<0.001 vs. mock-infected controls). Increased total CASP3-positive cells (+60%) and CASP3-positive cells in SOX2 (+70%), TBR2 (+65%), and CTIP2-positive cells (+60%) in ZIKV-infected organoids than mock-infected controls. Increased percentage of CASP3-positive cells in ZIKV-infected organoids at 5 (+20%, p<0.01 paired t-test vs. mock-infected controls) and 7 (+25%, p<0.0001 paired t-test vs. mock-infected controls) days post infection.

Wells MF Staining for Ki67 and CASP3. Bright field images and quantification of area per ZIKV MOI in AXL-WT and AXL-KO-Tm organoids.

Similar level of infection of Ki67-positive cells in AXL-WT and AXL-WT-Tm ZIKV-infected organoids. Similar decreases in area for AXl-WT and AXL-KO-Tm ZIKV-infected organoids at 72 (-25% for MOI 0.1, -35% for MOI 1 and 10) and 144 hours (-50% for MOI 0.1 and MOI 1, -100% for MOI 10) post-infection in comparison to mock-infected controls. Similar increases in CASP3-positive cells in AXL-WT and AXL-KO-Tm ZIKV-infected organoids comparison to controls (R-squared = 0.3815 for mock-infected and ZIKV-infected AXL-XO-Tm organoids vs. R=squared = -.3599 for mock-infected and ZIKV-infected AXL-WT organoids)

Xu M CASP3 assay with Emricasan treatment.

Increased percentage of CAS3-positive cells after ZIKV infection (+1.75%, p<0.001 one way ANOVA vs. mock-infected controls). No difference in percentage of CAS3-positive cells after ZIKV-infection when treated with 10 uM Emricasan.

Yoon KJ Staining for PAX6, EdU, PKC-lambda (adherens continuity), and Ki67.

Decreased percentage of EdU-positive cells and PAX6-EdU-positive cells in ZIKV-NS2A-expressed organoids (-17.5% for both, p<0.001 t-test vs. controls). Decreased percentage of PKC-lambda-positive cells in the ventricular surface in ZIKV-NS2A expressed organoids than controls (-60%, p<0.01 t-test vs. controls). Increased percentage of PAX6-positive cells with multipolar morphology in ZIKV-NS2A expressed organoids (+27.5%, p<0.01 t-test vs. controls). Decreased percentage of Ki67-positive cells in ZIKV-NS2A expressed organoids (-20%, p<0.01 t-test vs. controls).

Zhou T qRT-PCR analysis and bright field quantification of organoid area of ZIKV with mock-infected control and ZIKV and Hippeastrine hydrobromide treatment. Staining for Ki67, CASP3, SOX2, and TUJ1

Decreased area in ZIKV-infected organoids (-75%, p<0.001 unpaired two-tailed t-test vs. mock-infected controls). No difference between HH-treated organoids infected with ZIKV and mock-infected controls. More than 80% of SOX2-positive cells infected by ZIKV. SOX2-positive cells and TUJ1-positive cells are intermixed & lack discernible organization. Decreased percentage of Ki67-positive cells (-10%, p<0.05 unpaired two tailed t-test vs. mock-infected controls), but no difference between HH-treated ZIKV-infected organoids and mock-infected controls. Increased percentage of CASP3-positive cells (+20%, p<0.001 unpaired two tailed t-test vs. mock-infected controls), but no difference between HH-treated ZIKV-infected organoids and mock-infected controls.

32

Table 3. Judgment of the quality of evidence for intervention using OHAT Risk of Bias tool.Au

thor

Rand

omiz

ation

: W

as Z

IKV

expo

sure

ade

quat

ely

rand

omize

d?

(Hom

ogen

ous c

ell s

uspe

nsio

n =

no v

aria

tion)

Allo

catio

n Co

ncea

lmen

t: W

as th

e al

loca

tion

of Z

IKV

expo

sure

ad

equa

tely

con

ceal

ed to

rese

arch

ers?

(Ho

mog

enou

s sus

pens

ion

= no

var

iatio

n)

Expe

rimen

tal C

ondi

tions

: Wer

e th

e ex

perim

enta

l con

ditio

ns

iden

tical

acr

oss s

tudy

gro

ups?

(Sa

me

med

ia, s

olve

nt, i

ncub

ator

, pl

ate

cond

ition

s)

Blin

ding

dur

ing

Stud

y: W

ere

the

rese

arch

per

sonn

el b

linde

d to

th

e st

udy

grou

p du

ring

the

stud

y? (

Robo

tic sy

stem

s elim

inat

e ne

ed)

Inco

mpl

ete

Data

: W

ere

the

outc

ome

data

com

plet

e w

ithou

t att

rition

or e

xclu

sion

from

ana

lysi

s? (

Evid

ence

of w

ell o

r pla

te

loss

with

out e

xpla

natio

n)

Expo

sure

Cha

ract

eriza

tion:

Can

we

be co

nfide

nt in

the

expo

sure

ass

essm

ent?

(Pu

rity,

stab

ility

, sol

ubili

ty, v

olati

lity

of

subs

tanc

e)

Out

com

e As

sess

men

t: C

an w

e be

con

fiden

t in

outc

ome

asse

ssm

ent?

(Ac

cept

able

or w

ell e

stab

lishe

d m

etho

ds a

nd

blin

ding

unl

ess a

utom

ated

/no

hand

ling

betw

een

expe

rimen

ts

and

mea

sure

men

ts)

Repo

rting

: W

ere

all m

easu

red

outc

omes

repo

rted

?

Cugola FR Definitely low risk of bias- Homogenous cell suspension- ZIKV-BR, ZIKV-AF, YFV, Control

Definitely low risk of bias- Homogenous cell suspension- ZIKV-BR, ZIKV-AF, YFV, Control

Definitely low risk of bias- Same media, solvent, incubator, plate conditions

Definitely high risk of bias- No evidence of blinding to study group involving organoids, author indicated that investigators were not blinded to allocation during experiments and outcome assessment

Probably low risk of bias- Insufficient information provided about loss of organoids (all experiments used 3 organoids except for caspase-3 measurements, which used 10)

Definitely low risk of bias- Viral culture, amplification, and titration described (MOI = 0.1)

Definitely low risk of bias- Analysis of data performed by NIS Elements 3.22 software

Definitely low risk of bias- All measured outcomes outlined are recorded in main text or supplementary material

Dang J Definitely low risk of bias- Homogenous cell suspension- ZIKV, Control

Definitely low risk of bias- Homogenous cell suspension- ZIKV, Control

Definitely low risk of bias- Same media, solvent, incubator, plate conditions

Definitely high risk of bias- No evidence of blinding to study group involving organoids, author indicated that investigators were not blinded to allocation during experiments and outcome assessment

Probably high risk of bias (not recorded)- Insufficient information provided about number of organoids (only volume experiment recorded 5 organoids used)

Definitely low risk of bias- Viral culture, amplification, and titration described (MOI = 1)

Definitely low risk of bias- Analysis of data performed by Cytoscape with ClueGo add-on or ImageJ software

Definitely low risk of bias- All measured outcomes outlined are recorded in main text or supplementary material

Gabriel E Definitely low risk of bias

Definitely low risk of

Definitely low

Definitely high risk of

Probably high risk of

Definitely low risk of

Definitely low risk of

Definitely low risk of bias

33

- Homogenous cell suspension- ZIKV-AS, ZIKV-AM, ZIKV-AF, Control

bias- Homogenous cell suspension- ZIKV-AS, ZIKV-AM, ZIKV-AF, Control

risk of bias- Same media, solvent, incubator, plate conditions

bias- No evidence of blinding to study group involving organoids, author indicated that investigators were not blinded to allocation during experiments and outcome assessment

bias- Evidence of all organoids used was documented by table S2**Cortical plate thickness measured only for ZIKV-AS, not ZIKV-AM or ZIKV-AF**Values not shown for TUNEL assay**Organoid diameter measured for ZIKV-AF only, not ZIKV-AS or ZIKV-AM**Percentage of ventricular zone lumen exhibiting TUJ1-positive cells measured for ZIKV-AS and ZIKV-AM, not ZIKV-AF

bias- Viral culture, amplification, and titration described (MOI = 1)

bias- Analysis of data performed by ImageJ software

- All measured outcomes outlined are recorded in main text or supplementary material

Garcez P Definitely low risk of bias- Homogenous cell suspension- ZIKV, DENV2, control

Definitely low risk of bias- Homogenous cell suspension- ZIKV, DENV2, control

Definitely low risk of bias- Same media, solvent, incubator, plate conditions

Definitely high risk of bias- No evidence of blinding to study group involving organoids, author indicated that investigators were not blinded to allocation during experiments and outcome assessment

Probably low risk of bias- Insufficient information provided about number of organoids (3 or 6 organoids used)

Definitely low risk of bias- Viral culture, amplification, and titration described (MOI = 0.025 to 0.25)

Definitely low risk of bias- Analysis of data performed by Harmony 5.1 software

Definitely low risk of bias- All measured outcomes outlined are recorded in main text or supplementary material

Li C Definitely low risk of bias- Homogenous cell suspension- ZIKV, control

Definitely low risk of bias- Homogenous cell suspension- ZIKV, control

Definitely low risk of bias- Same media, solvent, incubator, plate conditio

Definitely high risk of bias- No evidence of blinding to study group involving organoids, author

Probably high risk of bias (not recorded)- Insufficient information provided about number of organoids

Definitely low risk of bias- Viral culture, amplification, and titration described (MOI = 1)

Definitely low risk of bias- Analysis of data performed by Imaris or ImageJ software

Definitely low risk of bias- All measured outcomes outlined are recorded in main text or supplementary material

34

ns indicated that investigators were not blinded to allocation during experiments and outcome assessment

(number of organoids not recorded)

Li Y Definitely low risk of bias- Homogenous cell suspension- ZIKV, control

Definitely low risk of bias- Homogenous cell suspension- ZIKV, control

Definitely low risk of bias- Same media, solvent, incubator, plate conditions

Definitely high risk of bias- No evidence of blinding to study group involving organoids, author indicated that investigators were not blinded to allocation during experiments and outcome assessment

Probably high risk of bias (not recorded)- Insufficient information provided about number of organoids (number of organoids not recorded)**Apoptosis measurements by CASP3 not quantified**Full statistical method not disclosed for each experiment

Definitely low risk of bias- Viral culture, amplification, and titration described (MOI = 1)

Definitely low risk of bias- Analysis of data performed by ImageJ software

Definitely low risk of bias- All measured outcomes outlined are recorded in main text or supplementary material

Qian X Definitely low risk of bias- Homogenous cell suspension- ZIKV, control

Definitely low risk of bias- Homogenous cell suspension- ZIKV, control

Definitely low risk of bias- Same media, solvent, incubator, plate conditions

Definitely low risk of bias- Investigators blinded to culture conditions

Probably high risk of bias- Insufficient information provided about number of organoids (3, 5, or 6 organoids used)**ZIKV-C strain limited to SOX2 and TBR2 staining**Statistical method not reported

Definitely low risk of bias- Viral culture, amplification, and titration described (1x or 0.25x dilution)

Definitely low risk of bias- Analysis of data performed by ImageJ software

Definitely low risk of bias- All measured outcomes outlined are recorded in main text or supplementary material

Sacramento CQ

Probably high risk of bias- Homogenous cell suspension- ZIKV, no controls

Probably high risk of bias- Homogenous cell suspension- ZIKV, no controls

Definitely low risk of bias- Same media, solvent, incubator, plate conditions

Definitely low risk of bias- Pharmacological assays kept blind

Probably high risk of bias (not recorded)- Insufficient information provided about number of organoids (number of

Definitely low risk of bias- Viral culture, amplification, and titration described (MOI = 1 or 10)

Definitely low risk of bias- Analysis of data performed by Harmony 5.1 software

Definitely low risk of bias- All measured outcomes outlined are recorded in main text or supplementary material (only ZIKV production/replication reported)

35

organoids not recorded)

Watanabe M

Definitely low risk of bias- Homogenous cell suspension- ZIKV, control

Definitely low risk of bias- Homogenous cell suspension- ZIKV, control

Definitely low risk of bias- Same media, solvent, incubator, plate conditions

Definitely high risk of bias- No evidence of blinding to study group involving organoids, author indicated that investigators were not blinded to allocation during experiments and outcome assessment

Definitely low risk of bias- Number of organoids provided for all experiments

Definitely low risk of bias- Viral culture, amplification, and titration described (MOI = 0.3 or 3.125)

Definitely low risk of bias- Analysis of data performed by ImageJ or ImageXpress software

Definitely low risk of bias- All measured outcomes outlined are recorded in main text or supplementary material

Wells MF Definitely low risk of bias- Homogenous cell suspension- ZIKV-PR, control

Definitely low risk of bias- Homogenous cell suspension- ZIKV-PR, control

Definitely low risk of bias- Same media, solvent, incubator, plate conditions

Definitely high risk of bias- No evidence of blinding to study group involving organoids, author indicated that investigators were not blinded to allocation during experiments and outcome assessment

Probably low risk of bias- Insufficient information provided about number of organoids (4 or 6 organoids used)**Ki67 measurements not quantified

Definitely low risk of bias- Viral culture, amplification, and titration described (MOI = 0.1, 1, or 10)

Definitely low risk of bias- Analysis of data performed by ImageJ software

Definitely low risk of bias- All measured outcomes outlined are recorded in main text or supplementary material

Xu M Definitely low risk of bias- Homogenous cell suspension- ZIKV, control

Definitely low risk of bias- Homogenous cell suspension- ZIKV, control

Definitely low risk of bias- Same media, solvent, incubator, plate conditions

Definitely high risk of bias- No evidence of blinding to study group involving organoids, author indicated that investigators were not blinded to allocation during experiments and outcome assessment

Probably low risk of bias- Insufficient information provided about number of organoids (6 organoids used)

Definitely low risk of bias- Viral culture, amplification, and titration described (MOI = 0.04 to 0.08)

Definitely low risk of bias- Analysis of data conducted in blinded fashion

Definitely low risk of bias- All measured outcomes outlined are recorded in main text or supplementary material

Yoon KJ Definitely low risk of bias- Homogenous cell suspension

Definitely low risk of bias- Homogenous cell

Definitely low risk of bias- Same media,

Definitely high risk of bias- No evidence of blinding to

Definitely low risk of bias- Number of organoids provided for

Definitely low risk of bias- Co-expression of GFP and

Definitely low risk of bias- Analysis of data performed

Definitely low risk of bias- All measured outcomes outlined are recorded in main text or

36

- Co-expression of GPF and ZIKV-NS2A or DENV-NS2A via electroporation, control

suspension- Co-expression of GPF and ZIKV-NS2A or DENV-NS2A via electroporation, control

solvent, incubator, plate conditions

study group involving organoids, author indicated that investigators were not blinded to allocation during experiments and outcome assessment

all experiments

ZIKV-NS2A or DENV-NS2A in ventricular radial glia cells in forebrain organoids by electroporation

by Zen or ImageJ software

supplementary material

Zhou T Definitely low risk of bias- Homogenous cell suspension- ZIKV, control

Definitely low risk of bias- Homogenous cell suspension- ZIKV, control

Definitely low risk of bias- Same media, solvent, incubator, plate conditions

Definitely high risk of bias- No evidence of blinding to study group involving organoids, author indicated that investigators were not blinded to allocation during experiments and outcome assessment

Definitely low risk of bias- 3 organoids provided for each experiments

Definitely low risk of bias- Viral culture, amplification, and titration described (5x10^5 PFU/ml)

Definitely low risk of bias- Analysis of data performed by ImageXpress software

Definitely low risk of bias- All measured outcomes outlined are recorded in main text or supplementary material

37

Table 4. Cellular effects of ZIKV according to lineage.

Stra

ins

Vira

l rep

licati

on a

fter 7

2 ho

urs

Pref

eren

tial i

nfec

tivity

Prog

enito

r cel

ls/ra

dial

glia

cells

(SO

X2+,

PAX

6+, N

estin

+, P

hosp

ho-v

imen

tin+)

Inte

rmed

iate

cells

(TBR

1+/T

BR2+

)

Prol

ifera

ting

cells

(Ki6

7+, p

hosp

ho-H

3)

Mat

ure

neur

ons

(CTI

P2+,

TU

J+)

Apop

tosis

(TUN

EL+,

CAS

P3+)

Viab

ility

(EdU

+)

Chan

ge in

org

anoi

d siz

e, a

rea,

vol

ume,

or

thic

knes

s

African (MR766)

↑22,23,27,31

Progenitor cells > Mature neurons 19,25

↓20,21,22,23

↓21

↓20,21,27,31

↓21

↑21,23,27,31

↓20

↓22,23,24,25,27

,31

Latin America(PRVABC59, Brazilian, FB-GWUH-2016)

↑19,23,25

Progenitor cells > Mature neurons19,21,26

↓21,23,28

↓21,28

↓21

↓21,23,28

↑21,23,28,29

N/A ↓21,23,28,29

Asian(FSS13025, H/PF/2013)

↑23

Progenitor cells > Mature neurons21,25

↓23

N/A N/A ↓23

↑ 23,30

N/A ↓23

38