ANALYSIS OF THE GROWTH OF SICKLE HEMOGLOBIN FIBERS OBSERVED
BY DIC MICROSCOPY
Sickle hemoglobin is a point mutation of the b chain in the hemoglobin tetramer (a2b2) that allows the hemoglobin molecules to form long, 14 stranded fibers. These fibers distort and rigidify the cell, thereby causing the pathology of sickle cell disease. The elongation of such fibers can be visualized by differential interference contrast (DIC) microscopy using visible light. First employed in this system by Briehl and coworkers (Nature 1990 345: 833-835, JMB 1995 245:710-723), it has also been used by Vekilov and coworkers (JMB 2007 365: 425-39), and most recently Castle, Odde and Wood (Sci Adv 2019 5:eaau1086) who made a detailed analysis of growth rates and their fluctuations and who have generously shared their data with us. Electron microscopy studies that have produced a fiber model have been based on extensive averaging of images; no cryo-EM has yet been done. There is some structural evidence that molecules can assume more than one arrangement in the fiber lattice which might lead to a large variation of elongation rates, as reported by Castle et al. When we analyze the data we find a Poisson distribution of elongation rates, and this distribution accounts for the observed variance originally thought to be anomalously large. In such a case, the variance is directly related to the mean and not an independent measurement. We hypothesize that this distribution is the result of a single pixel in the image containing a large number of elongation steps. Such an issue could arise in any DIC analysis of an elongating protein polymer.
Sickle hemoglobin is a point mutation of the b chain in the hemoglobin tetramer (a2b2) that allows the hemoglobin molecules to form long, 14 stranded fibers. These fibers distort and rigidify the cell, thereby causing the pathology of sickle cell disease. The elongation of such fibers can be visualized by differential interference contrast (DIC) microscopy using visible light. First employed in this system by Briehl and coworkers (Nature 1990 345: 833-835, JMB 1995 245:710-723), it has also been used by Vekilov and coworkers (JMB 2007 365: 425-39), and most recently Castle, Odde and Wood (Sci Adv 2019 5:eaau1086) who made a detailed analysis of growth rates and their fluctuations and who have generously shared their data with us. Electron microscopy studies that have produced a fiber model have been based on extensive averaging of images; no cryo-EM has yet been done. There is some structural evidence that molecules can assume more than one arrangement in the fiber lattice which might lead to a large variation of elongation rates, as reported by Castle et al. When we analyze the data we find a Poisson distribution of elongation rates, and this distribution accounts for the observed variance originally thought to be anomalously large. In such a case, the variance is directly related to the mean and not an independent measurement. We hypothesize that this distribution is the result of a single pixel in the image containing a large number of elongation steps. Such an issue could arise in any DIC analysis of an elongating protein polymer.