Editorial

Zika: Path to pathogenicity

Scott B. Halstead

Zika virus is a member of the Flavivirus genus, related genetically and structurally to the four dengue viruses. Zika is maintained in complex African zoonoses, and at some time in the past spilled into the Aedes aegypti as part of humans’ urban transmission cycle. Because many flaviviruses circulate in African populations, it was logical to wonder if these antibodies enhanced Zika infections in Fc-receptor–bearing cells and thereby contributed to maintenance of the virus in nature. Despite a tentative answer of “yes” obtained from in vitro studies, those humans infected with Zika in Africa had mild disease, and no change in the severity or diversity of human Zika clinical disease has been reported since the virus was first recovered in 1947.

Furthermore, there were no unusual Zika virus (ZIKV) infections reported in India or Southeast Asia, where ZIKV and dengue viruses (DENV) were found to be co-circulating as early as 1953. Without warning during the past decade, large outbreaks occurred on Pacific islands revealing a marked change in Zika pathogenicity. In Tahiti, from 2013 to 2014, an outbreak of ZIKV was followed in 4 weeks by an epidemic of Guillain-Barré syndrome. Of the 42 cases of Guillain-Barré syndrome, 95% had evidence of prior DENV infection, although this percentage did not differ from controls. In Brazil, ZIKV was found to infect and cross the placenta in pregnant women and then destroy fetal tissues. Next, ZIKV was found to infect the male reproductive tract resulting in sexual transmission.

While the genetic diversity of ZIKV strains suggested the possibility that mutations can explain this new pathogenicity, there is also a growing interest in the possibility that antibody-dependent enhancement (ADE) of ZIKV infections by DENV antibodies play a role. In this hypothesis, DENV antibodies that enhance ZIKV infections result in high viremia levels that pass into the placenta or increase phenomena that otherwise occur at very low rates, such as Guillain-Barré syndrome and sexual transmission. Two studies in particular have shown that polyclonal antibodies from patients experiencing secondary DENV infections enhanced or neutralized ZIKV. Derived monoclonal antibodies (mAb) to the fusion loop epitope failed to neutralize but reliably enhanced ZIKV in Fc-receptor–bearing K-562 human myelogenous leukemia or the U-937 human monocytic cell line. On the other hand, DENV mAb directed at a conformational quaternary epitope formed at the interface of two envelope protein monomers (EDE1) potently neutralized ZIKV in a picomolar range similar to that of DENV. X-ray crystallographic structures of antigen-binding fragments of EDE1 and EDE2 in complex with the ZIKV envelope protein have been obtained.

But, speculations from in vitro data are insufficient to establish ZIKV ADE by DENV antibodies as a human pathogenic process. DENV ADE results from the targeting of myeloid cells during human infections. In A129 mice lacking type I interferon receptors, peripheral infection with ZIKV produced observable sickness and high titers of virus in the spleen, liver and brain, but no attempt was made to identify target cells. Epidemiological studies were essential to link severe DENV infections to secondary heterotypic DENV infections or primary DENV infections in infants who circulate passively acquired polyvalent DENV immunoglobulin G antibodies. If Zika infection intensity is controlled by DENV antibody, ADE disease outcomes could be influenced by the complicated immunological preconditions that control the severity of DENV infections. For example, might the antibodies to a single prior DENV infection be enhancing ZIKV while antibodies from two or more prior DENV infections protect against it? This is what structural and enhancement studies on monoclonal DENV antibodies from a single infection and neutralization by mAb obtained after two infections suggest. Could Zika enhancement vary by antibodies to different DENV types? What difference might the length of the interval between DENV and ZIKV infections play? Finally, a gene abundant in sub-Saharan African populations appears to moderate severe secondary DENV infections. Might this or similar genes moderate Zika pathogenicity? Expertly planned investigations are needed quickly as the current ZIKV epidemic may soon recede.

Disclosure: Halstead reports no relevant financial disclosures.

Scott B. Halstead

Zika virus is a member of the Flavivirus genus, related genetically and structurally to the four dengue viruses. Zika is maintained in complex African zoonoses, and at some time in the past spilled into the Aedes aegypti as part of humans’ urban transmission cycle. Because many flaviviruses circulate in African populations, it was logical to wonder if these antibodies enhanced Zika infections in Fc-receptor–bearing cells and thereby contributed to maintenance of the virus in nature. Despite a tentative answer of “yes” obtained from in vitro studies, those humans infected with Zika in Africa had mild disease, and no change in the severity or diversity of human Zika clinical disease has been reported since the virus was first recovered in 1947.

Furthermore, there were no unusual Zika virus (ZIKV) infections reported in India or Southeast Asia, where ZIKV and dengue viruses (DENV) were found to be co-circulating as early as 1953. Without warning during the past decade, large outbreaks occurred on Pacific islands revealing a marked change in Zika pathogenicity. In Tahiti, from 2013 to 2014, an outbreak of ZIKV was followed in 4 weeks by an epidemic of Guillain-Barré syndrome. Of the 42 cases of Guillain-Barré syndrome, 95% had evidence of prior DENV infection, although this percentage did not differ from controls. In Brazil, ZIKV was found to infect and cross the placenta in pregnant women and then destroy fetal tissues. Next, ZIKV was found to infect the male reproductive tract resulting in sexual transmission.

While the genetic diversity of ZIKV strains suggested the possibility that mutations can explain this new pathogenicity, there is also a growing interest in the possibility that antibody-dependent enhancement (ADE) of ZIKV infections by DENV antibodies play a role. In this hypothesis, DENV antibodies that enhance ZIKV infections result in high viremia levels that pass into the placenta or increase phenomena that otherwise occur at very low rates, such as Guillain-Barré syndrome and sexual transmission. Two studies in particular have shown that polyclonal antibodies from patients experiencing secondary DENV infections enhanced or neutralized ZIKV. Derived monoclonal antibodies (mAb) to the fusion loop epitope failed to neutralize but reliably enhanced ZIKV in Fc-receptor–bearing K-562 human myelogenous leukemia or the U-937 human monocytic cell line. On the other hand, DENV mAb directed at a conformational quaternary epitope formed at the interface of two envelope protein monomers (EDE1) potently neutralized ZIKV in a picomolar range similar to that of DENV. X-ray crystallographic structures of antigen-binding fragments of EDE1 and EDE2 in complex with the ZIKV envelope protein have been obtained.

But, speculations from in vitro data are insufficient to establish ZIKV ADE by DENV antibodies as a human pathogenic process. DENV ADE results from the targeting of myeloid cells during human infections. In A129 mice lacking type I interferon receptors, peripheral infection with ZIKV produced observable sickness and high titers of virus in the spleen, liver and brain, but no attempt was made to identify target cells. Epidemiological studies were essential to link severe DENV infections to secondary heterotypic DENV infections or primary DENV infections in infants who circulate passively acquired polyvalent DENV immunoglobulin G antibodies. If Zika infection intensity is controlled by DENV antibody, ADE disease outcomes could be influenced by the complicated immunological preconditions that control the severity of DENV infections. For example, might the antibodies to a single prior DENV infection be enhancing ZIKV while antibodies from two or more prior DENV infections protect against it? This is what structural and enhancement studies on monoclonal DENV antibodies from a single infection and neutralization by mAb obtained after two infections suggest. Could Zika enhancement vary by antibodies to different DENV types? What difference might the length of the interval between DENV and ZIKV infections play? Finally, a gene abundant in sub-Saharan African populations appears to moderate severe secondary DENV infections. Might this or similar genes moderate Zika pathogenicity? Expertly planned investigations are needed quickly as the current ZIKV epidemic may soon recede.

PAGE BREAK

Disclosure: Halstead reports no relevant financial disclosures.