Sie sind auf Seite 1von 3


Rapid repurposing of drugs for COVID-19

The emergence of a new coronaviral respiratory disease calls for repurposing existing drugs

By R. Kiplin Guy1, Robert S. DiPaola2, Frank relatively high mortality, filling the gap for (AZ), a widely used broad-spectrum antibi-
Romanelli1, Rebecca E. Dutch2 coronavirus-specific drugs is urgent. otic, also blocks autophagosome clearance in
The coronavirus life cycle (see the figure) human cells (11) and replication of the Zika

n late fall 2019, a novel acute respira- involves a number of potentially targetable virus and influenza virus in human cells in
tory disease, called coronavirus disease steps, including endocytic entry into host vitro (12). Preliminary results from a small
2019 (COVID-19) emerged in Wuhan, cells [involving angiotensin-converting en- randomized trial of HCQ in COVID-19 pa-
China. COVID-19 is caused by severe zyme 2 (ACE2) and transmembrane prote- tients report a reduction in time to clinical
acute respiratory syndrome–corona- ase serine 2 (TMPRSS2)], RNA replication resolution (13). A small open-label trial has
virus 2 (SARS-CoV-2) (1, 2). COVID-19 and transcription [involving helicase and demonstrated increased reduction in viral
has been declared a pandemic by the RNA-dependent RNA polymerase (RdRp)], load for COVID-19 patients receiving the
World Health Organization and continues translation and proteolytic processing of combination of HCQ and AZ relative to HCQ
to spread across the globe. Most patients viral proteins (involving chymotrypsin-like alone, although this study has been heavily
recover within 1 to 3 weeks. However, a and papain-like proteases), virion assem- criticized because of post hoc removal of

Downloaded from on June 5, 2020

small proportion (~5%) develop severe ill- bly, and release of new viruses through the several subjects from the study analysis (14).
ness that can progress to acute respiratory exocytic systems (4). In addition to virally These hypothesis-generating studies have
distress syndrome (ARDS), which can lead encoded targets, numerous host targets are justified emergency approval of their use for
to death. Currently, only supportive care is essential for viral replication and disease COVID-19 in the United States, where they
available; patients would greatly benefit progression (3). are both being widely used.
from the availability of direct therapeu- The cellular receptor for SARS-CoV-2 However, both HCQ and AZ have po-
tic approaches. One approach to identify- is ACE2 (5). Recombinant human ACE2 tential cardiac toxicity (QT prolongation,
ing therapeutics is to repurpose approved (rhACE2, or APN01) is currently under de- which can lead to fatal arrhythmia), and
drugs developed for other uses, which velopment as a treatment for acute lung HCQ additionally has the potential for
takes advantage of existing detailed infor- injury and pulmonary arterial hypertension negative effects on the eye. Understanding
mation on human pharmacology and toxi- and has proven well tolerated in a phase risk-benefit ratios is paramount if these
cology to enable rapid clinical trials and 1 trial in healthy volunteers. rhACE2 has drugs are to become a standard of care for
regulatory review. been shown to significantly reduce viral en- COVID-19. Several post hoc analyses car-
The coronaviruses are single-stranded try into human cell–derived organoids (6), ried out in the United States and Europe
RNA viruses that infect vertebrates and presumably by acting as a decoy for virus suggest modest benefit, at best, from HCQ
move between different host species (3). binding. This has lent support to the clini- monotherapy for COVID-19 patients; one
With the emergence of SARS-CoV-2, there cal trials that are investigating blockade of large post hoc analysis among U.S. veterans
are now seven coronaviruses that are known viral entry with APN01 for COVID-19 pa- suggests that there is harm to patients from
to infect humans. Four of them (HCoV- tients. Successful viral entry requires pro- HCQ. Given the mechanistic rationale but
229E, HCoV-OC43, HCoV-NL63, and HCoV- teolytic processing of the viral coat spike lack of well-designed clinical studies and
HKU1) are responsible for ~30% of cases of glycoprotein (S), which can be carried out potential for drug-induced toxicity, there is
the common cold in humans. Two of them by TMPRSS2 (7). The TMPRSS2 inhibitor a key need for controlled, randomized trials
caused recent epidemics that had consid- camostat (7) is approved in Japan for the to test the efficacy and safety of these drugs
erable associated mortality: SARS-CoV-1, treatment of chronic pancreatitis and post- for COVID-19 patients.
which emerged in 2002–2003 and causes operative gastric reflux and is generally well After uncoating, the viral genomic RNA is
~10% mortality, and Middle East respira- tolerated, although rare serious side effects used for cap-dependent translation to pro-
tory syndrome coronavirus (MERS-CoV), have been reported. Both camostat and the duce two polypeptides, which are then auto-
which emerged in 2012, is still active, and related agent nafamostat (8) block SARS- proteolytically processed to produce several
causes ~35% mortality. Both epidemics af- CoV-2 replication in TMPRSS2-expressing viral proteins, including RdRp and two pro-
fected a relatively small number of patients human cells. Camostat has been shown to teases. Although the proteases might seem
compared with COVID-19, which is more block infection with SARS-CoV-2 in a mouse attractive targets given the number of viral
transmissible for several reasons, including model. Therefore, there is a strong rationale protease inhibitors previously developed for
asymptomatic carriers, long latency period, to support clinical trials with these drugs HIV and other viruses, they are only distantly
and high infectivity. Before COVID-19, only for COVID-19, which have already been ini- related to other viral proteases. The combina-
SARS-CoV-1 and MERS-CoV caused severe tiated in the Netherlands and Germany. tion of the HIV protease inhibitors lopinavir
disease. Therefore, coronaviral drug discov- Coronaviruses use the endolysosomal and ritonavir (15) proved clinically ineffec-
ery has been a small effort relative to that pathway to enter the cell before uncoating. tive for COVID-19 patients, as had previously
for other viral diseases such as influenza. Chloroquine (CQ) and hydroxychloroquine been the case for the same combination in
Given the rapid spread of COVID-19 and its (HCQ) are antimalarial drugs that affect SARS-CoV-1 disease. Therefore, further re-
endosomal function and block autophago- purposing with this class of drugs is poorly
some-lysosome fusion (9). Both drugs have justified—although there are other protease
College of Pharmacy, University of Kentucky, Lexington,
KY, USA. 2College of Medicine, University of Kentucky, been shown to inhibit SARS-CoV-2 replica- inhibitors in early-stage drug discovery that
Lexington, KY, USA. Email: tion in cellular models (8, 10). Azithromycin are directed to the coronavirus proteases.

SCIENCE 22 MAY 2020 • VOL 368 ISSUE 6493 829

Published by AAAS

Production of the replication complex pro- als). Additionally, phenotypic screening individual patients during epidemic peaks
teins, including the helicase and RdRp, al- approaches are being developed on the ba- on the basis of clinical studies that involve
lows for genomic replication of the virus and sis of either viral entry or replication that small numbers of patients with ensuring
for production of subgenomic RNAs, which could be used to survey approved drugs and that well-designed, randomized clinical tri-
are also translated to produce structural drug candidates much more widely. Both of als are carried out rapidly to provide proof
and coat proteins. The helicase is theoreti- these approaches may widen the available that they are safe and efficacious. COVID-19
cally an attractive target, but it is divergent classes of drugs for consideration. is expected to be active permanently, and
from other viral helicases, and there is no The key issue with any of these potential several seasons of disease peaks are likely
evidence that the herpes simplex virus heli- treatments is to balance the oppositional before herd (population) immunity is estab-
case inhibitors amenamevir or pretelivir are needs of making treatment decisions for lished. The difficulty is to coordinate rapid
effective against coronaviruses. hypothesis-generating studies
RdRp carries out both replica- during this first peak to jus-
tion and transcription of the vi- Possible targets in the coronavirus life cycle tify a smaller number of well-
ral RNA, making it a clear target This simplified coronavirus life cycle shows the processes and proteins that controlled large trials to be ex-
for blocking the viral life cycle. could be therapeutically targeted with existing drugs that have the potential to be ecuted in later peaks to pro-
Because RdRp is a critical pro- repurposed for the treatment of COVID-19. vide the data needed for ap-
tein for many viruses, a number proval of drugs for COVID-19.
of broad-spectrum RdRp inhibi- Researchers, ethics boards, and
tors are either approved or in 1 Attachment regulators are accustomed to de-
clinical trials, including remde- and entry veloping trial plans over months,
sivir and favipiravir. Remdesivir not weeks—a time frame that is

Downloaded from on June 5, 2020

was initially developed to treat TMPRSS2 ACE2 not afforded during this emer-
the filoviruses that cause Ebola Cell gent situation. It is necessary for
and Marburg diseases and has Camostat, rhACE2 all involved to work faster and
proven safe in trials during nafamostat Endocytosis more efficiently and then posi-
the past two Ebola epidem- tion the well-justified drugs for
ics. However, it is less effective Endosome registration-enabling trials dur-
for Ebola than antibody-based ing the next peak. j
treatments that prevent the vi-
rus from entering human cells. 2 Uncoating RE FERENCES AND NOT ES
gRNA 1. C. del Rio, P. N. Malani, JAMA 323, 1339
Remdesivir was subsequently hydroxychloroquine,
shown to be active against both azithromycin
2. N. Zhu et al., N. Engl. J. Med. 382, 727
SARS-CoV-1 and MERS-CoV (2020).
3. T. S. Fung, D. X. Liu, Annu. Rev. Microbiol.
in animal models. Favipiravir 73, 529 (2019).
was developed for influenza
3 gRNA replication Transcription
4. G. Li, E. De Clercq, Nat. Rev. Drug Discov.
and approved in Japan in 2014, 19, 149 (2020).
specifically for new pandemic RdRp 5. J. Lan et al., Nature 10.1038/s41586-
Polypeptide 020-2180-5 (2020).
influenza outbreaks. Both rem- gRNA
chains 6. V. Monteil et al., Cell 10.1016/j.
desivir and favipiravir are active Remdesivir, gRNA cell.2020.04.004 (2020).
against SARS-CoV-2 in human 7. M. Hoffmann et al., Cell 181, 271 (2020).
Favipiravir Ribosomes sgRNAs 8. M. Wang et al., Cell Res. 30, 269 (2020).
cells in vitro (7). Remdesivir has 9. M. Mauthe et al., Autophagy 14, 1435
been rapidly advanced into sev- (2018).
eral clinical trials for COVID-19, 4 Translation 10. J. Liu et al., Cell Discov. 6, 16 (2020).
in ER and Golgi 11. M. Renna et al., J. Clin. Invest. 121, 3554
and early informal data being (2011).
released from those trials sug- Replicated Nonstructural Structural and 12. D. H. Tran et al., J. Antibiot. 72, 759
gest that remdesivir is effective, gRNA proteins coat proteins (2019).
13. Z. Chen et al., medRxiv,
but such datasets need to be 2020.03.22.20040758 [Preprint]. 10
used cautiously for generaliz- April 2020.
ing the understanding of either
14. P. Gautret et al., Int. J. Antimicrob.
safety or efficacy. Further ran- Agents 10.1016/j.ijantimi-
domized, controlled trials with 5 Assembly cag.2020.105949 (2020).
RdRp inhibitors are justified 15. B. Cao et al., N. Engl. J. Med. 10.1056/
and needed. NEJMoa2001282 (2020).
The best justified drugs for ACKNOWL EDGMENTS
repurposing to treat COVID-19 The authors acknowledge the participation
patients are the host-factor– of the COVID-19 Unified Research Experts
targeted drugs HCQ, AZ, and (CURE) Alliance team at the University of

Kentucky in discussions underlying this

camostat and nafamostat and Golgi piece. We have no funding supporting the
the viral RdRp–targeted drugs ER writing of this article and no conflicts of
remdesivir and favipiravir. A interest. The therapeutics discussed in this
number of other drugs are also 6 Virion release Perspective are undergoing clinical testing
and are not currently approved for the treat-
being considered, although COVID-19, coronavirus disease 2019; ER, endoplasmic reticulum; gRNA, genomic RNA; RdRp, ment of COVID-19.
RNA- dependent RNA polymerase; rhACE2, recombinant human angiotensin-converting enzyme 2;
with less supporting evidence SARS-CoV-2, severe acute respiratory syndrome–coronavirus 2; sgRNA, subgenomic RNA; TMPRSS2,
Published online 8 May 2020
(see supplementary materi- transmembrane protease serine 2. 10.1126/science.abb9332

830 22 MAY 2020 • VOL 368 ISSUE 6493 SCIENCE

Published by AAAS
Rapid repurposing of drugs for COVID-19
R. Kiplin Guy, Robert S. DiPaola, Frank Romanelli and Rebecca E. Dutch

Science 368 (6493), 829-830.

DOI: 10.1126/science.abb9332originally published online May 8, 2020

Downloaded from on June 5, 2020



REFERENCES This article cites 10 articles, 0 of which you can access for free


Use of this article is subject to the Terms of Service

Science (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement of
Science, 1200 New York Avenue NW, Washington, DC 20005. The title Science is a registered trademark of AAAS.
Copyright © 2020 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of
Science. No claim to original U.S. Government Works