Chloroquine is a potent inhibitor of SARS coronavirus infection and spread

In late two thousand two, a novel respiratory illness appeared in Guangdong Province, China. Within months, severe acute respiratory syndrome, or SARS, had reached at least thirty countries. The causative agent, a new coronavirus called SARS-CoV, was unlike any previously known coronavirus. And when it emerged, there were no proven prophylactic or post-exposure therapies. None. Laboratories scrambled. Small interfering RNA, passive antibody transfer, DNA vaccination, interferons, monoclonal antibodies targeting the spike protein—all were under investigation. But none were ready for rapid deployment. The situation demanded something faster: an existing drug with a known safety profile that could be repurposed and tested immediately. Chloroquine fit that description. First identified in nineteen thirty-four, it is a weak base that has been used for decades against malaria, amoebiasis, and autoimmune diseases. Its cellular behavior is well understood. When chloroquine enters a cell, it diffuses into acidic organelles—endosomes, lysosomes, Golgi vesicles—becomes protonated, and gets trapped there, raising the pH of those compartments. That pH elevation matters enormously for viruses like SARS-CoV, which depend on acidic endosomal conditions to fuse with host membranes and release their genetic cargo into the cell. Disrupt that acidification, and you can stop entry cold. But Vincent and colleagues, working at the Centers for Disease Control, identified a second mechanism that makes chloroquine's story more interesting. To understand it, you need to know how SARS-CoV gets into a cell in the first place. The virus's spike glycoprotein—a type I membrane protein studding the viral surface—binds to a host receptor called angiotensin-converting enzyme 2, or ACE2. That binding initiates uptake. Now, ACE2 isn't just a simple protein sitting on the cell surface. It's heavily decorated with sugar chains—glycans—added in the endoplasmic reticulum and then modified further in the Golgi apparatus. Those terminal sugar modifications can influence how tightly a ligand, like a viral spike protein, actually binds. Chloroquine, by raising Golgi pH, interferes with that terminal glycosylation process. The team showed this biochemically. Using metabolic labeling and immunoprecipitation, they identified two forms of ACE2 in Vero E6 cells—African green monkey kidney cells, a standard model for coronavirus research. One form migrated at around one hundred five kilodaltons, sensitive to an enzyme called endoglycosidase H, which marks it as the immature form still in the endoplasmic reticulum. The other form migrated at around one hundred thirteen kilodaltons, endoglycosidase H resistant—the Golgi-modified, mature form. When cells were treated with chloroquine, the one hundred thirteen-kilodalton form shifted to a faster migration, consistent with loss of terminal glycan modifications. At twenty-five micromolar chloroquine, the Golgi form resolved similarly to ACE2 treated with neuraminidase, an enzyme that specifically strips terminal sialic acid residues. So chloroquine was essentially mimicking the removal of those terminal sugars. Critically, flow cytometry—a technique that measures protein abundance on the cell surface by labeling live cells with antibodies—showed that ten micromolar chloroquine did not significantly reduce the total amount of ACE2 at the cell surface. The receptor was still there. It was just less well-dressed. Under-glycosylated ACE2, Vincent and colleagues argue, may have reduced binding affinity for the viral spike, potentially slowing or preventing the initiation of infection. That's the prophylactic picture. Now here's where the experiments get particularly compelling. The team treated Vero E6 cell monolayers with chloroquine for twenty to twenty-four hours before exposing them to SARS-CoV. After infection, they used indirect immunofluorescence—a method where a primary antibody recognizes viral antigens and a fluorescently labeled secondary antibody makes them visible under a microscope—to count antigen-positive cells. The results were striking. At point one micromolar chloroquine, infected cells dropped by twenty-eight percent. At one micromolar, by fifty-three percent. At ten micromolar, infection was completely abolished. No detectable viral antigen. This was reproduced across three independent experiments. Ten micromolar. That's a concentration well within the range of what chloroquine achieves in human plasma during standard malaria treatment—reported as one point six to twelve point five micromolar. The drug was hitting its antiviral target at doses people were already taking safely. But pre-treatment is only half the story. The therapeutic question—can you give this drug after someone is already infected and still limit viral spread—is the harder and more clinically relevant one. And that's where things get particularly interesting. Vincent and colleagues infected cells first, let the virus adsorb for one hour, then washed it away and added chloroquine. Even in this post-infection setting, the drug was dramatically effective. As little as point one to one micromolar reduced infection by fifty percent. Maximum inhibition of roughly ninety to ninety-four percent was observed with thirty-three to one hundred micromolar. The calculated half-maximal inhibitory concentration—the ED50—was four point four micromolar, plus or minus one. And the drug remained significantly effective even when added three or five hours after initial viral adsorption, though higher concentrations were needed at those later time points. This is the therapeutic window. The virus has already attached. Entry has begun. And chloroquine can still halt the spread. To understand why, the team compared chloroquine to ammonium chloride—another compound that raises endosomal pH by a different chemical route. Ammonium chloride at concentrations of five millimolar or higher produced ninety-three to ninety-nine percent inhibition of SARS-CoV, both before and after infection. The parallel between chloroquine and ammonium chloride's timing and potency points strongly toward endosomal pH elevation as the dominant mechanism in the post-infection setting. When a virus is inside an endosome trying to fuse with the membrane, and you suddenly raise the pH, you subvert that fusion event. The virus is trapped. It can't escape into the cytoplasm to replicate. The team also examined the viral spike protein itself to ask whether chloroquine was disrupting its biosynthesis or processing. Using pulse-chase labeling with radioactive cysteine, they tracked the spike precursor—a one hundred ninety-kilodalton protein—as it was processed into smaller mature forms. At antiviral concentrations of ten and fifty micromolar chloroquine, there was no substantial impact on spike biosynthesis, Golgi modification, or oligomeric assembly. Only at one hundred micromolar did biosynthesis suffer. Ammonium chloride, by contrast, blocked the appearance of Golgi-modified spike forms at ten millimolar and above. This difference is important. It tells us the two drugs share a pH-elevating mechanism but aren't identical in their effects on the viral machinery—and that at clinically relevant chloroquine concentrations, the spike protein itself is not the primary target. Putting it together: when chloroquine is given before exposure, it produces under-glycosylated ACE2 at the cell surface that may bind the spike protein less efficiently, slowing the initiation of infection. When given after exposure, it raises endosomal pH and disrupts the pH-dependent fusion steps the virus needs to complete entry. Two mechanisms, two windows of activity, one drug. The finding that these effects occur at one to ten micromolar—concentrations achievable in clinical dosing—is what makes this paper more than a basic science curiosity. Vincent and colleagues are explicit about it: chloroquine has a long history of human use without significant detrimental side effects. These results were also independently supported by Keyaerts and colleagues, who reported antiviral effects of chloroquine against SARS-CoV using quantitative reverse transcription polymerase chain reaction, providing corroboration from a different detection method. What the paper cannot tell us—and the authors are honest about this—is whether these cell-culture findings translate to living organisms. All the data are from Vero E6 monolayers. The team explicitly calls for animal model studies and more detailed binding assays examining how chloroquine-induced ACE2 glycosylation changes actually affect spike-receptor interaction at the molecular level. Those are the logical next steps: confirm the binding effect directly, validate efficacy in vivo, and define the dosing regimen that maps the one to ten micromolar effective window onto real clinical practice. What this paper did, in two thousand five, was establish the scientific foundation clearly and rigorously. A cheap, widely available, well-tolerated drug could block SARS-CoV in a primate cell system—both before and after exposure—at concentrations humans could safely achieve. That's a concrete, evidence-based starting point. From cell to clinic is never a short road. But Vincent and colleagues showed where the road begins. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.