Renovating Hemoglobin

Using nuclear magnetic resonance spectroscopy as a tool for structural analysis, Ho has over many years of work gained insight into hemoglobin cooperativity at the atomic level -- understanding what parts of the structure shift and change as successive oxygen atoms bind to the protein. In the last few years, he overcame one of the major stumbling blocks to a hemoglobin-based blood supply: He developed a method -- based in DNA-cloning techniques -- that can produce large quantities of human hemoglobin in laboratory bacterial cultures. Now he is focusing on hemoglobin's cooperativity and a closely related property -- its oxygen affinity.

When oxygenated hemoglobin reaches its delivery point in human tissue, it has to let go of its payload -- in other words, it has to have low oxygen affinity. In red blood cells, a built-in mechanism adjusts hemoglobin's oxygen affinity, but in laboratory-synthesized hemoglobin, outside the cell environment, this mechanism doesn't exist.

To compensate, Ho wants to engineer hemoglobin that has the appropriate oxygen affinity. Each hemoglobin subunit, like any protein, is a chain of amino acids, and the sequence of these basic units determines how the protein functions. With simulations on the C90, Ho and Madrid are testing what happens if they change one of the amino acids in a subunit.

Their simulations to date have focused on the regions where subunit chains connect with each other. Several naturally occurring variants of normal hemoglobin exist in which a particular site at the subunit interface -- known as beta 99 -- has mutated to a different amino acid. These mutant hemoglobins have abnormally low cooperativity and high oxygen affinity -- just the opposite of what is desired. "Can we restore normal function by an additional mutation, by genetic engineering?" asks Ho. The answer, they have learned, is yes.

Interface Region of Computer-Designed Hemoglobin
Two subunits (blue and light blue) are shown with their associated hemes (red). In this abnormal hemoglobin, the amino acid at the beta-99 site, normally aspartic acid, is mutated to asparagine (yellow), which eliminates hydrogen bonds between the subunits, resulting in lowered cooperativity and high oxygen affinity. Changing one other amino acid -- tyrosine, alpha-42 to aspartic acid (green) -- creates new hydrogen bonds (dashed lines) with asparagine and arginine, beta-40 (gray), which partially restores cooperativity and oxygen affinity.

The simulations indicated that changing an amino acid in the neighborhood of the mutated site would restore or create hydrogen bonds between the subunits. These bonds are normally present in deoxy-hemoglobin, but don't exist in the mutants. Ho surmised that restoring these bonds would restore normal cooperativity and oxygen affinity.

Guided by the C90 simulations, Ho turned to the laboratory. "We have had two excellent successes," says Ho, "two mutants that look very promising." One of them restores cooperativity to one of the abnormal hemoglobins. Another is a bioengineered hemoglobin with low oxygen affinity, a big step toward the goal of a viable blood substitute.

With these successes, Ho's research strategy has proven itself, and Ho plans to tackle other problems in engineering a hemoglobin blood substitute: "We need to have some kind of rational basis for making mutants in the laboratory. We can't just make anything that comes to our head. Supercomputing gives some guidance, and it also can tell us the detailed mechanisms involved, precise information that experiments can't give. Modeling alone doesn't give the whole picture; we need to test the ideas. But the combination of computer modeling and experiments -- this is a good marriage."

go back to the main screen