Read with caution!

This post was written during early stages of trying to understand a complex scientific problem, and we didn't get everything right. The original author no longer endorses the content of this post. It is being left online for historical reasons, but read at your own risk.

On our visit to NIH back in April we learned a great deal about stem cell assays and I’ve been meaning to blog about it ever since.  Suppose that you wanted to create a stem cell line from an adult, differentiate the stem cells into a plausible simile of brain tissue, and then conduct a high-throughput screen for therapeutics.  There are all sorts of good questions this would raise, perhaps most important of which is choosing the right assay (how will you tell if a drug worked?), and we don’t have all the answers to those questions yet, but what we did learn at NIH was a very clear picture of the steps, timeline and costs involved in such an assay.  The numbers.

Here is a basic listing of the steps involved, with a timeline for each.

  1. Obtain cells.  Preferably fibroblasts (i.e. skin), ideally 2 million cells (about a half inch square skin punch) but no fewer than half a million.  Blood (20mL) also works but isn’t ideal; same with bone marrow (too painful?).
  2. Expand cells in culture.  7 – 21 days.
  3. Induce pluripotence.  Introduce transcription factors to transform the adult cells into iPSC, minimizing off-target effects. Use a non-integrating vector– for blood, sendai virus; for fibroblasts, plasmids.  Both are commercially available.  28 days.
  4. Test and expand the line.  Produce a bankable stock of iPSC.  150 days.
  5. Testing for stability, etc.  60 days.
  6. Make neural stem cells (NSCs) from iPSC.  14 days.
  7. Expand the neural stem cell line.  You can’t do this after full differentiation because neurons are post-mitotic.  45 days.
  8. Differentiate the NSCs into neurons.  12 days.
  9. Prepare for the high-throughput screen.  12-15 days.
  10. Conduct the screen.  3 days.
  11. Analyze the data.  7 days.

Total: about a year.

In developing the screen, you have your choice of plate size: 96-, 384- or 1536-well plates, but the volume of each well gets smaller so that the total number of cells is about 1-10 million cells per plate in any case.  It takes about 2M iPSCs to make 10-100M NSCs, which in turn can make 10-100M neurons.  (Hence all the line-expanding steps above).

If you want to screen only the already FDA-approved drugs, there are on the order of 3,000 of them.  If you want to do a complete discovery screen, you could test 100 – 300K compounds from the NIH compound library.  The cost for a high-throughput screen (and I think this is just drug and materials costs for the screening stage itself) is about $1/well, and screens are usually run in triplicate, so you’re looking at about $3 per compound you would like to screen.  (Multiply that again by the number of stem cell lines you want to screen against if you have positive and negative controls for example).  Making and banking the iPSC will also cost $25K/line and converting to NSC will cost $10-$15K/line.

In terms of discovery screens, it’s important to bear in mind that any new compound (or even the most minor modification to an existing FDA-approved drug) has to go through the entire FDA approval process.  This takes on average 14 years for normal drugs, average 7 years for fast track drugs.  The all-time record fastest approval has been 3 years.  Of course, FDA approval follows human clinical trials, so naturally some patients will have access to a drug before it is approved, particularly for rare and orphan diseases where an emergency / “compassionate use” exemption is possible.

The right partnerships will be crucial at multiple stages in the process. Here are some potential partners to consider or reach out to:

  • Tissue collection, i.e. the clinical process of seeing a patient and collecting a tissue sample.  The NIH clinical center itself does this but it takes a long time; New York Stem Cell Foundation (located in upper upper upper Manhattan) also does this.  There are also commercial options available.
  • Making NSCs from iPSCs.  Baylor (Houston) and City of Hope (Los Angeles) have the most experienced labs; UW-Madison, UC-Davis, Yale and NIH itself also do this.
  • Screening.  Only a handful of companies/schools/institutions in the U.S. do high-throughput screening.  Of note are: Galena Biopharma (Oregon), Scripps (SoCal), Yale, Stanford, Broad Institute (Cambridge), NCATS (Rockville MD).

So those are some of the key points and numbers pertaining to stem cell assays.  But most importantly: you need a postdoc to do all the work :)