244 ‒ The history of the cell, cell therapy, gene therapy, and more | Siddhartha Mukherjee

KEYPOINTS

• 01

  Cell biology advanced steadily, while genetics stalled for decades.

  Mukherjee contrasts the two histories. After Mendel's paper, genetics had roughly 40 years of silence. Cell biology grew steadily from the 17th-century microscope to Rudolf Virchow's 19th-century claim that every function, and every illness, comes down to how cells behave. Attia says the early figures impressed him even more than the genome decoders because they had so few tools.

• 02

  Germ theory and the cell powered the leap from 'Medicine 1.0' to 2.0.

  Attia frames the shift as two processes: a new way of thinking (the scientific method) and new tools for seeing, such as microscopes that revealed germs. Mukherjee agrees that seeing microbes and linking them to disease changed medicine. He adds a clinical lesson: the most important question is often what the patient thinks is wrong.

• 03

  Multicellularity may evolve more easily than once assumed.

  Single-celled organisms remain wildly successful, living even in boiling water and volcanic vents. So the puzzle is why multicellular life exists at all. William Ratcliff and his advisor Michael Travisano repeatedly selected yeast that settled fastest in a tube. Within about 30–40 cycles, perhaps 60–90 days, clusters of cells emerged that stayed together, a new form of life evolved by simple selection. Mukherjee also notes size limits set by the ratio of volume to surface area.

• 04

  There are four ways to use cells as medicine.

  First, drugs that change a cell's behavior, such as antibiotics killing microbes but sparing human cells. Second, transferring unmodified cells, as in blood transfusion. Third, using cells as factories for products like insulin or antibodies such as Herceptin. Fourth, genetically modifying cells and putting them into the body. Mukherjee stresses that 'gene therapy is really cell therapy': the right gene in the wrong cell is useless or dangerous.

• 05

  CAR-T cells can cure some blood cancers, but cost is a structural problem.

  T cells are taken from a patient, engineered with a virus to target cancer (for example, CD19 on B-cell cancers) and infused back. They work in lymphoma, leukemia and myeloma, but not yet well in solid tumors. US costs run about $500,000 to $1 million per patient. Mukherjee's team in India has treated about 11–12 patients with very good early data and aims for 20–50-fold lower costs. He links US prices partly to the roughly 90% of drugs that fail, which companies cite to justify pricing; he calls that argument somewhat specious.

• 06

  CRISPR is like editing one word in a vast library.

  Picture the genome as about 80,000 books. Old viral gene therapy slipped a new page in, sometimes into the wrong book. CRISPR, from the work of Jennifer Doudna, Emmanuelle Charpentier, Feng Zhang and others, can go to one book, one page and one word and change 'verbal' to 'herbal.' Mukherjee's group is editing bone marrow, and he discusses gene editing for sickle cell disease, which is caused by a single amino-acid change.

• 07

  Gene editing crosses an ethical line when it moves from disease to desire.

  Mukherjee recounts He Jiankui editing CCR5 in embryos so that twins would be HIV-resistant. Because they were conceived by IVF with washed sperm, their risk of getting HIV from their father was essentially zero. He says the change served the desire to be first, not to treat disease, and made unintended, imprecise germline changes. Editing for Huntington's (100% penetrant) sits clearly on the disease side. Less penetrant genes such as APOE4 or Lp(a), and traits like height, move toward desire.

• 08

  Most mental illness is 'death by a thousand cuts,' which makes it hard to edit.

  A few rare, highly penetrant variants run in families. He mentions a book about a family with several children with schizophrenia. But like height, most risk for depression, schizophrenia and autism comes from many small genetic nudges, many not yet identified. Current editing handles only one to three genes.

• 09

  Science has two kinds of anomaly: an 'eye in the sandstorm' and 'sand in the eye.'

  In the first, nothing fits, as with the move from Newtonian to quantum physics. In the second, a theory is almost right but one fact won't fit. His example is the gap between neurons, the synapse: why convert electrical signals to chemical ones and back? The answer was weighting. Adjustable synaptic weights make learning networks possible, like artificial neural networks, rather than a 'box of wires.'