James Lowenthal’s first blog post for AST 214: How many virus particles to infect?

One thing I’ve always wanted to know: how many virus particles does it take to infect a person?  One?  A million?  If “it depends,” what does it depend on?

This seems important to know especially now, as we focus almost all our global attention on the COVID19 pandemic and doing our best to contain or limit its human and economic harm.  Are we adopting the right policies to avoid infection?  People are sewing masks in their living rooms to donate to hospitals and health care workers; is this the right approach?  Many are fearful of touching the mail, or of entering a grocery store; are those fears well-founded scientifically?  Is the advice broadcast by governments (local, state, national, international) based on reliable scientific information?

Here are a few interesting things I’ve learned recently about people and viruses (with apologies to you biologists):

  • A human body has about 1013 human cells but also 1014 (100 trillion) other organisms representing thousands of species: bacteria, fungi, and protozoans, most of them benign and beneficial and symbiotic.  Each of us is outnumbered 10-to-1 on our own body!
    Source: https://www.ncbi.nlm.nih.gov/books/NBK26917/#A4614
  • A virus is mostly a tiny snippet of DNA  — the instructions to replicate itself — wrapped in a little bubble of fat (lipid).  No nucleus, no mitochondria, no production of new proteins the way cells usually function.  A virus sneaks its way into a regular cell, then brainwashes that cell into copying the virus’s DNA many times over (thousands to millions) until it burst open, spewing out new versions of the virus.
  • It seems the way the tests work to detect the coronavirus that causes COVID19 is by processing a small sample taken say from a patient’s nose or throat and putting it through an acceleration process (polymerase chain reaction, or PCR) that amplifies many times over the presence of the virus’s DNA.  If only a few cycles (n<29) of that process produce a detectable result,  there must have been a large amount of the virus’ DNA to begin with, a large “viral loading”.  If many cycles are required (n>40), there must have been only a small amount.  The number of cycles required is referred to as “cycle threshold” or Ct.  Here’s a letter showing the viral loading of the coronavirus over time in patients in Wuhan, China: https://www.nejm.org/doi/full/10.1056/NEJMc2001737

    The higher the viral loading, the more the person is “shedding” viruses, and therefor contagious.  It seems that infected people are most contagious a few days after onset of symptoms, and that contagiousness drops steadily after that.  But there’s a pretty wide variation and scatter, and the trends don’t seem super conclusive yet.  I’m sure with the much higher numbers coming, the trends in contagiousness will be better understood.
  • Here’s one article in the Washington Post by a professor of exposure and assessment science at Harvard saying, basically, “Social distancing and hand-washing: good!  Panic about infection from deliveries or grocery shopping: not necessary.”

After all that — I still don’t know how many virus particles it takes to infect a human!  Please let me know if you find out — thanks!