Stress-induced unfolded protein response contributes to Zika virus–associated microcephaly


Accumulating evidence support a causal link between Zika virus (ZIKV) infection during gestation and congenital microcephaly. However, the mechanism of ZIKV-associated microcephaly remains unclear. We combined analyses of ZIKV-infected human fetuses, cultured human neural stem cells and mouse embryos to understand how ZIKV induces microcephaly. We show that ZIKV triggers endoplasmic reticulum stress and unfolded protein response in the cerebral cortex of infected postmortem human fetuses as well as in cultured human neural stem cells. After intracerebral and intraplacental inoculation of ZIKV in mouse embryos, we show that it triggers endoplasmic reticulum stress in embryonic brains in vivo. This perturbs a physiological unfolded protein response within cortical progenitors that controls neurogenesis. Thus, ZIKV-infected progenitors generate fewer projection neurons that eventually settle in the cerebral cortex, whereupon sustained endoplasmic reticulum stress leads to apoptosis. Furthermore, we demonstrate that administration of pharmacological inhibitors of unfolded protein response counteracts these pathophysiological mechanisms and prevents microcephaly in ZIKV-infected mouse embryos. Such defects are specific to ZIKV, as they are not observed upon intraplacental injection of other related flaviviruses in mice.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Fig. 1: ZIKV induces ER stress and UPR in human cortices in vivo and in hNSCs in vitro.
Fig. 2: ICV injection of ZIKV induces microcephaly in mouse embryos.
Fig. 3: ICV injection of ZIKV induces activation of ER stress and UPR in mouse embryonic brains.
Fig. 4: ICV injection of ZIKV promotes direct neurogenesis at the expense of indirect neurogenesis in mouse in vivo.
Fig. 5: PERKi rescues the neurogenic balance upon ICV injection of ZIKV in mouse in vivo.
Fig. 6: ICV and intraplacental (IPL) injection of ZIKV, but not yellow fever (YFV) or WNV viruses, induces microcephaly and cell death in mouse brains.


  1. 1.

    Gupta, A., Tsai, L. H. & Wynshaw-Boris, A. Life is a journey: a genetic look at neocortical development. Nat. Rev. Genet. 3, 342–355 (2002).

  2. 2.

    Rash, B. G. & Grove, E. A. Area and layer patterning in the developing cerebral cortex. Curr. Opin. Neurobiol. 16, 25–34 (2006).

  3. 3.

    Laguesse, S., Peyre, E. & Nguyen, L. Progenitor genealogy in the developing cerebral cortex. Cell Tissue Res. 359, 17–32 (2015).

  4. 4.

    Sarnat, H. B. & Flores-Sarnat, L. A new classification of malformations of the nervous system: an integration of morphological and molecular genetic criteria as patterns of genetic expression. Eur. J. Paediatr. Neurol. 5, 57–64 (2001).

  5. 5.

    Sarnat, H. B. & Flores-Sarnat, L. Neuroembryology and brain malformations: an overview. Handb. Clin. Neurol. 111, 117–128 (2013).

  6. 6.

    Woods, C. G. & Parker, A. Investigating microcephaly. Arch. Dis. Child. 98, 707–713 (2013).

  7. 7.

    Cauchemez, S. et al. Association between Zika virus and microcephaly in French Polynesia, 2013-15: a retrospective study. Lancet 387, 2125–2132 (2016).

  8. 8.

    Driggers, R. W. et al. Zika virus infection with prolonged maternal viremia and fetal brain abnormalities. N. Engl. J. Med. 374, 2142–2151 (2016).

  9. 9.

    Mlakar, J. et al. Zika virus associated with microcephaly. N. Engl. J. Med. 374, 951–958 (2016).

  10. 10.

    Cugola, F. R. et al. The Brazilian Zika virus strain causes birth defects in experimental models. Nature 534, 267–271 (2016).

  11. 11.

    Miner, J. J. et al. Zika virus infection during pregnancy in mice causes placental damage and fetal demise. Cell 165, 1081–1091 (2016).

  12. 12.

    Li, C. et al. Zika virus disrupts neural progenitor development and leads to microcephaly in mice. Cell Stem Cell 19, 120–126 (2016).

  13. 13.

    Onorati, M. et al. Zika virus disrupts phospho-TBK1 localization and mitosis in human neuroepithelial stem cells and radial glia. Cell Rep. 16, 2576–2592 (2016).

  14. 14.

    Qian, X. et al. Brain-region-specific organoids using mini-bioreactors for modeling ZIKV exposure. Cell 165, 1238–1254 (2016).

  15. 15.

    Tang, H. et al. Zika virus infects human cortical neural progenitors and attenuates their growth. Cell Stem Cell 18, 587–590 (2016).

  16. 16.

    Wu, K. Y. et al. Vertical transmission of Zika virus targeting the radial glial cells affects cortex development of offspring mice. Cell Res. 26, 645–654 (2016).

  17. 17.

    Laguesse, S. et al. A dynamic unfolded protein response contributes to the control of cortical neurogenesis. Dev. Cell 35, 553–567 (2015).

  18. 18.

    Chavali, P. L. et al. Neurodevelopmental protein Musashi-1 interacts with the Zika genome and promotes viral replication. Science 357, 83–88 (2017).

  19. 19.

    Bell, T. M., Field, E. J. & Narang, H. K. Zika virus infection of the central nervous system of mice. Arch. Gesamte. Virusforsch. 35, 183–193 (1971).

  20. 20.

    Hamel, R. et al. Biology of Zika virus infection in human skin cells. J. Virol. 89, 8880–8896 (2015).

  21. 21.

    Monel, B. et al. Zika virus induces massive cytoplasmic vacuolization and paraptosis-like death in infected cells. EMBO. J. 36, 1653–1668 (2017).

  22. 22.

    Paul, D. & Bartenschlager, R. Flaviviridae replication organelles: oh, what a tangled web we weave. Annu. Rev. Virol. 2, 289–310 (2015).

  23. 23.

    Pillai, P. S. et al. Mx1 reveals innate pathways to antiviral resistance and lethal influenza disease. Science 352, 463–466 (2016).

  24. 24.

    Staeheli, P., Haller, O., Boll, W., Lindenmann, J. & Weissmann, C. Mx protein: constitutive expression in 3T3 cells transformed with cloned Mx cDNA confers selective resistance to influenza virus. Cell 44, 147–158 (1986).

  25. 25.

    Borgs, L. et al. Dopaminergic neurons differentiating from LRRK2 G2019S induced pluripotent stem cells show early neuritic branching defects. Sci. Rep. 6, 33377 (2016).

  26. 26.

    Bernard-Marissal, N. et al. Reduced calreticulin levels link endoplasmic reticulum stress and Fas-triggered cell death in motoneurons vulnerable to ALS. J. Neurosci. 32, 4901–4912 (2012).

  27. 27.

    Olgar, Y., Ozdemir, S. & Turan, B. Induction of endoplasmic reticulum stress and changes in expression levels of Zn(2 + )-transporters in hypertrophic rat heart. Mol. Cell. Biochem. (2017).

  28. 28.

    Sessa, A. et al. Tbr2-positive intermediate (basal) neuronal progenitors safeguard cerebral cortex expansion by controlling amplification of pallial glutamatergic neurons and attraction of subpallial GABAergic interneurons. Genes Dev. 24, 1816–1826 (2010).

  29. 29.

    Axten, J. M. et al. Discovery of GSK2656157: an optimized PERK inhibitor selected for preclinical development. ACS Med. Chem. Lett. 4, 964–968 (2013).

  30. 30.

    Walter, P. & Ron, D. The unfolded protein response: from stress pathway to homeostatic regulation. Science 334, 1081–1086 (2011).

  31. 31.

    Gupta, A. et al. NCOA3 coactivator is a transcriptional target of XBP1 and regulates PERK-eIF2α-ATF4 signalling in breast cancer. Oncogene 35, 5860–5871 (2016).

  32. 32.

    Tsuru, A., Imai, Y., Saito, M. & Kohno, K. Novel mechanism of enhancing IRE1α-XBP1 signalling via the PERK-ATF4 pathway. Sci. Rep. 6, 24217 (2016).

  33. 33.

    Bonnin, A. et al. A transient placental source of serotonin for the fetal forebrain. Nature 472, 347–350 (2011).

  34. 34.

    Ghouzzi, V. E. et al. 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 Dis. 7, e2440 (2016).

  35. 35.

    Yoon, K. J. et al. Zika-virus-encoded NS2A disrupts mammalian cortical neurogenesis by degrading adherens junction proteins. Cell Stem Cell 21, 349–358.e6 (2017).

  36. 36.

    Brault, J. B. et al. Comparative analysis between flaviviruses reveals specific neural stem cell tropism for Zika virus in the mouse developing neocortex. EBioMedicine 10, 71–76 (2016).

  37. 37.

    Lambert, N. et al. Genes expressed in specific areas of the human fetal cerebral cortex display distinct patterns of evolution. PLoS One 6, e17753 (2011).

  38. 38.

    Cau, E., Gradwohl, G., Fode, C. & Guillemot, F. Mash1 activates a cascade of bHLH regulators in olfactory neuron progenitors. Development 124, 1611–1621 (1997).

  39. 39.

    Gladwyn-Ng, I. E. et al. Bacurd2 is a novel interacting partner to Rnd2 which controls radial migration within the developing mammalian cerebral cortex. Neural Dev. 10, 9 (2015).

  40. 40.

    Gladwyn-Ng, I. et al. Bacurd1/Kctd13 and Bacurd2/Tnfaip1 are interacting partners to Rnd proteins which influence the long-term positioning and dendritic maturation of cerebral cortical neurons. Neural Dev. 11, 7 (2016).

  41. 41.

    Lanciotti, R. S. et al. Genetic and serologic properties of Zika virus associated with an epidemic, Yap State, Micronesia, 2007. Emerg. Infect. Dis. 14, 1232–1239 (2008).

  42. 42.

    Tielens, S. et al. Elongator controls cortical interneuron migration by regulating actomyosin dynamics. Cell Res. 26, 1131–1148 (2016).

Download references


The authors are thankful for technical help from the GIGA-Imaging Platform of ULg; M. Leruez-Ville for human sample collection; P.V. Drion, E.D. Valentin and C. Grignet for the flaviviral facility; C. d’Alessandro, M. Sambon and M. Lavina for technical assistance; E. Simon-Loriere for sharing ZIKV quantification protocol; and E. Peyre for graphical assistance. L.N., I.G.-N. and C.C. receive financial support from F.R.S.-F.N.R.S. This work was supported by the European Union’s Horizon 2020 Research and Innovation Programme under ZIKAlliance Grant Agreement N° 734548 (to L.N. and M.L.) and by the European Virus Archives goes Global (EVAg) project under grant agreement N° 653316. M.L. is also funded by Institut Pasteur, Inserm and LabEx IBEID. L.N. and P.V. are funded by F.R.S.-F.N.R.S., the Fonds Léon Fredericq, the Fondation Médicale Reine Elisabeth, the Fondation Simone et Pierre Clerdent and the Belgian Science Policy (IAP-VII network P7/20). L.N. is funded by ARC (ARC11/16-01) and the ERANET Neuron STEM-MCD; P.V. is funded by the WELBIO Program of the Walloon Region, the AXA Research Fund, the Fondation ULB, ERANET Neuron Microkin, ERANET E-rare Euromicro and ERC-2013 ERC-2013-AG-340020.

Author information




I.G.-N., L.C.B., C.A., C.C., T.C., M.L. and L.N. designed the study. I.G.-N. set up animal models for ZIKV infection and, together with T.C., generated and analyzed in vivo data with help of G.M., L.C.B., M.A., C.A. and C.C. N.T. and M.T. performed and interpreted TEM analyses. L.C.B. generated data with hNSCs and performed analyses with help of C.A., I.G.-N. and C.C. C.A. and M.A. analyzed human brain samples provided by B.B., F.E.-R., M.B., I.S. and P.V. M.F. shared antibodies. L.N. contributed to data interpretation and wrote the manuscript with help from M.L. and input from all coauthors.

Corresponding authors

Correspondence to Marc Lecuit or Laurent Nguyen.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Integrated Supplementary Information

Supplementary Figure 1 ZIKV infection triggers ER stress in human cortices and hNSCs.

a, Model of cortical neurogenesis related to UPR. b, Venn diagram, and top regulated genes common to three models studying microcephaly upon ZIKV infection or activation of ER stress. c-d, Expression pattern of ZIKV mRNAs in the developing cerebral cortex of an ZIKV-infected fetus (ZIKV#1, 25 GW) detected by antisense (AS) probes but not sense probes (S) c. Uninfected brain (WT, 21GW) does not show ZIKV mRNA signal d (n = 3 independent experiments). e-f, Human neural stem cells (hNSCs) analyzed by qRT-PCR (n = 4 biologically independent samples) to detect OCT4 ( P = 0.0286, U = 0), NESTIN ( P = 0.0286, U = 0), PAX6 ( P = 0.0286, U = 0), and SOX1 ( P = 0.0286, U = 0) e. Immunolabelings of PAX6 (grey), SOX1 (red) and NESTIN (green) (n = 3 biologically independent samples) f. g, Boxplots showing increased expression of ATF3 ( P < 0.05, U = 0), CHAC1 ( P < 0.05, U = 0), CHOP ( P < 0.05, U = 0), and PDI ( P < 0.05, U = 0) upon tunicamycin treatment (TM) (n = 3 biologically independent samples). e-g, Data are presented as boxplots of median ± first to third quartiles while whiskers extend to maxima and minima with each symbol representing one biologically independent sample (n = 3 to 4) as indicated on boxplots and analyzed by either two-sided e or one-sided g Mann-Whitney test. Scale bars represent 50 µm f. P < 0.05.

Supplementary Figure 2 Activation of PERK and IRE-1 pathways in ZIKV infected mouse brains from WT or Ifnr-knockout mice induce ER stress and microcephaly.

a, Intracerebroventricular (ICV) co-injection of ZIKV and PERKi at E12.5 rescues microcephaly of newborn pups. Data are presented as histograms and error bars of mean ± SEM respectively and analyzed by one-way ANOVA followed by Bonferonni post hoc comparison test to compare brain weight (F2,43=59.06, P < 0.0001), normalized cortical length (F2,43=65.85, P < 0.0001), and cortical width (F2,43=74.12, P < 0.0001) of pups after Mock, ZIKV or ZIKV + PERKi as indicated on the x-axes (14, 14, 16 brains from separate litters respectively) a. b-d, Histograms and related immunolabelings showing that ICV ZIKV injection into E12.5 brains increases number of Pax6+ ( P < 0.05, U = 0) but not Tbr2+ (P = 0.150, U = 1.5) cells in E18.5 cortices (n = 7 to 9 animals); the averaged Pax6-intensity remains unchanged (P = 0.9172, U = 20) between conditions (n = 96 to 106 cells from 3 independent embryonic brains) b-c. Immunolabelings showing Dapi (blue), ZIKV (red) Pax6 (green), and Tbr2 (grey) (n = 7 to 9 independent embryonic brains) d. Data are presented as boxplots of median ± first to third quartiles while whiskers extend to maxima and minima with each symbol representing independent embryonic brains (n = 3) as indicated and analyzed by two-sided Mann-Whitney test (b-d). e-f, Western blot analyses of proteins extracted from E15.5 mouse brains ICV injected with Mock or ZIKV at E12.5. Representative immunoblots showing increased phosphorylation of PERK ( P < 0.05, U = 0) and IRE-1in ZIKV-infected brains and quantification of phosphorylated/unphosphorylated protein ratio after normalization of individual samples (n = 2 to 3 independent embryonic brains) with respect to corresponding conditions e. Immunoblot and corresponding scatter-plots showing increased phosphorylation of eIF2a protein (n = 2 to 3 independent embryonic brains, P < 0.05, U = 0) f. Data are presented as scatter-plots of means ± SEM with each symbol representing biologically independent samples as indicated on and analyzed by two-sided Mann-Whitney test e-f. g, Corresponding histograms representing fold change of UPR factor transcripts in E14.5 brains from an Ifnr−/− dam intraperitoneally injected with ZIKV at gestational day 9.5. Microdissected cortices were analyzed by qRT-PCR to detect PDI, Atf4, Atf5, Chac1, and Chop; one representative experiment for each condition. Scale bar represents 100 µm d. P < 0.001, and P < 0.05.

Supplementary Figure 3 Comparison of the expression pattern of ZIKV using two types of antibodies and proliferation assessment.

a-b,Immunolabelings of brain slices from E14.5 embryos infected by ZIKV after intracerebroventricular injection at E12.5 and followed by in utero electroporation of GFP at E13.5. Immunolabelings show Tbr2 (blue), Sox2 (light grey), GFP (green), and ZIKV (red, using two types of antibodies, as indicated in a and b (n = 3 independent experiments from 3 different embryonic brains). c-d, Assessment of proliferation rate (Ki67+, white) of GFP+ apical progenitors (green) in Mock condition or after ZIKV infection (red). Nuclei are counterstained with Dapi (blue) (n = 3 to 6 independent experiments) c-d. Data is presented as boxplot of median ± first to third quartiles while whiskers extend to maxima and minima with each symbol representing independent embryonic brains (n = 3 to 6, P < 0.116, U = 3), as indicated on boxplots and analyzed by two-sided Mann-Whitney test c. Scale bar represent 200 µm a, b, 100 µm (boxed regions in a and b), or 50 µm d. ns means not significant.

Supplementary Figure 4 ZIKV does not impair cell survival after 2 d but leads to PERKi-sensitive ER stress.

a, Data presented as histograms and error bars of mean ± SEM respectively and analyzed by one-way ANOVA followed by Bonferonni post hoc comparison test, representing cycling progenitors that express Ki67 and expressed as percentage of control, experimental conditions as mentioned on the graph (n = 11 independent embryonic E14.5 brains per condition infected at E12.5) a. b-c, ZIKV induces non-cell autonomous changes of the neurogenic balance in E14.5 brains after infection at E12.5 (n = 4 to 5 independent embryonic brains, Sox+Tbr2+, F9,56=8.717, P < 0.0001; Tbr2+Tbr1+, F9,56=8.883, P < 0.0001) b. Immunolabelings of brain slices from E14.5 embryos showing Dapi (blue), ac-caspase 3 (gray), GFP (green), and ZIKV (red) and related data showing the percentage of cells per bin that express either ZIKV, ac-caspase 3, or co-express the combination of markers as indicated (n = 4 to 5 independent embryonic brains) c. Data presented as histograms and error bars of mean ± SEM respectively and analyzed by two-way ANOVA followed by Bonferonni post hoc comparison test. d-e, Immunolabelings on E14.5 cortical slices from 8 ZIKV-infected E14.5 brain showing ZIKV (red), βIII-tubulin (blue) and ac-caspase 3 (green) (n = 3 independent embryonic brains) (d, areas show ZIKV-infected (white arrow) and uninfected (open arrow dying neurons). Immunolabelings on E18.5 cortical slices from Mock or ZIKV-infected E18.5 brain showing ZIKV (red), PDI (green), and Dapi (blue) (n = 3 independent embryonic brains) e. f-h, treatment with PERKi reduces markers for ER stress in E18.5 brains after ICV injection at E12.5. Immunolabelings of cortical sections from Mock, ZIKV, or ZIKV+PERKi injected brains (5, 7 and 3 brains from separate litters per condition, respectively) to detect Dapi (blue), ZIKV (red), Calnexin (green) (f, F2,14=11.00, P < 0.0001), and Dapi (blue), ZIKV (red), Calreticulin (green) (g, F2,14=12.02, P < 0.0001) (n = 3 to 7 independent experiments). qRT-PCR performed on total RNA extract from E18.5 micro-dissected cortices after ICV injection of ZIKV at E12.5 to detect the following ER stress or UPR actors: Atf4 (F3,23=8.319, P=0.0399), Atf5 (F3,24=8.008, P=0.0458), and PDI (F3,22=6.888, P=0.0756) h. Data are presented as boxplots of median ± first to third quartiles while whiskers extend to maxima and minima with each symbol representing independent embryonic brains (n = 3 to 7) as indicated, and analyzed by Kruskal-Wallis followed by Dunn’s post hoc comparison test f-h. i, Data are presented as boxplots of median ± first to third quartiles while whiskers extend to maxima and minima with each symbol representing independent embryonic brains (n = 6 to 9) as indicated, indicating the percentage of cortical cells infected by ZIKV (P = 0.9172, U = 20) and the average ZIKV-intensity (P = 0.6870, t = 0.4115, df = 14), and analyzed by either two-sided Mann-Whitney or unpaired two-tailed Student’s t test respectively, in infected cells from ZIKV and ZIKV+PERKi injected brains i. j, Data is presented as boxplots of median ± first to third quartiles while whiskers extend to maxima and minima with each symbol representing independent embryonic brains (n = 6 to 7 brains from separate litters) as indicated, and analyzed by unpaired two-tailed Student’s t test, indicating the number of viral copies after in utero co-injection of either ZIKV, ZIKV+PERKi (P = 0.1141, t = 1.716, df = 11) or ZIKV+IRE-1i ( P = 0.0364, t = 2.382, df = 14) in brain extracts j. P < 0.001, P < 0.01, P < 0.05; ns means not significant. Scale bars are 50μm d, e.

Supplementary Figure 5 ZIKV triggers apoptosis via induction of terminal UPR and expression of Chop.

a-b, Data showing number of Ctip2+ (F2,19=2.750, P=0.2614) a and Cux1+ (F2,19=7.665, P=0.0140) b cortical neurons in E18.5 brains after treatment with inhibitors as indicated, and presented as boxplots of median ± first to third quartiles while whiskers extend to maxima and minima with each symbol representing one independent sample (n = 4 to 8) as indicated on boxplots, and analyzed by Kruskal-Wallis followed by Dunn post-hoc tests. c, Representative images of immunolabeled coronal sections from brains of E18.5 embryos infected by ZIKV at E12.5 treated by co-injection of DMSO or PERKi with detection of Dapi (blue), ZIKV (red), PDI (green) (n = 5 independent embryonic brains per condition). d-e, Suppression of Chop, a terminal downstream target of PERK pathway, improves survival of ZIKV-infected cortical cells. Boxplot demonstrating Chop siRNA suppresses Chop expression in transiently transfected Neuro2A cells followed by qRT-PCR (n = 4 independent experiment per condition, F4,20=16.53, P=0.0024) d. Boxplot demonstrating the comparison of the percentages of ac-caspase 3+ cells in E16.5 mouse brains, after ICV injection of either ZIKV or ZIKV+PERKi at E12.5 followed by in utero co-electroporation of either Chop or control siRNA with pCAGGS-GFP (n = 5 embryonic brains per condition, F5,34 = 24.69, P = 0.0002) e. Data presented as boxplots of median ± first to third quartiles while whiskers extend to maxima and minima with each symbol representing one biologically independent sample as indicated on boxplots, and analyzed by Kruskal-Wallis followed by Dunn post hoc comparison test. f, Representative images of immunolabeled coronal sections from brains of E18.5 embryos infected by ZIKV at E12.5. ZIKV+control siRNA or ZIKV+Chop siRNA E18.5 brains with detection of Dapi (blue), ZIKV (red), GFP (green) and ac-caspase 3. g, Schematic representation of the mechanisms triggered by ZIKV in the developing cortex. By targeting the APs, ZIKV generates ER stress that raises UPR, leading to an overall reduction of the neuronal output as a result of increased direct neurogenesis (red Arrow). Infected projection neurons suffer from chronic ER stress that activates the UPR target Chop to promote apoptosis (target). Single in vivo injection of PERKi at the time of infection partially releases both pathophysiological events and prevents microcephlay in mouse embryos. Scale bar represent 100 µm c, or 50 µm f. P < 0.01, and P < 0.05; ns means not significant. ROI = region of interest.

Supplementary Figure 6 Full scans of the western blots presented in Supplementary Fig. 2.

Full scans of the blots immunolabeled for eIF2a and pEIF2a a, PERK, pPERK, IRE1a and pIRE1a b, and ß-Actin (c and d controls of a and b, respectively). Each blotting membrane has been cut in several fragments to perform independent immunodetections. These fragments have been manually assembled to reconstitute the original membrane before the scanning.

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Gladwyn-Ng, I., Cordón-Barris, L., Alfano, C. et al. Stress-induced unfolded protein response contributes to Zika virus–associated microcephaly. Nat Neurosci 21, 63–71 (2018).

Download citation

Further reading