Amifostine
The Search of the Mechanism of Radioprotection by Aminothiols
Following completion of a postdoctoral fellowship at NRC I accepted a position at the Defence Research Board to work on protection against radiation. The project assigned to me was to explain how certain aminothiols protect against ionizing radiation; an easy task to define but difficult to execute. Aminothiols are compounds with the following basic structure: NH2RSH; where R represents a carbon chain of variable length.
My first step was to try to determine the chemical processes involved when such compounds are exposed to ionizing radiation. A puzzling aspect of radioprotection was that some compounds were protective, for example cysteamine (NH2CH2CH2SH), whereas other similar compounds such a penicillamine HOOCCH(NH2)C(CH3)2SH, were completely without protective properties. It had already been established that only compounds with an unbranched chain of two or three carbons between the sulfur and nitrogen atoms were protective. Cysteine was chosen as an example of a protective aminothiol. In water, it oxidises readily to a disulfide so that was used for radiation chemistry studies. The products obtained by irradiation of the disulfide in water with gamma rays were identified. Next, penicillamine was investigated using the same technique as an example of an aminothiol that did not protect against radiation. However, the same types of products were found in both cases. Even when a mixed disulfide of cysteine and penicillamine was used, there was no significant difference in the range of products between the two components.
The next approach was to use pulse radiolysis; a new technique that had just become available. I collaborated with Hugh Gillis and Norman Klassen at the National Research Council of Canada (NRC) to study the very early processes in the radiation chemistry of the sulfur-containing compounds. The linear accelerator (LINAC) at NRC had been adapted to measure the spectra of very short-lived species such as free radicals produced by radiation pulses of 10 nano seconds duration. Adams and Michael at Mount Vernon Hospital in UK had used this technique to study the disulfide cystamine and had identified an anion-radical (RSSR-) which had a strong absorption band in the UV. We used the pulse radiation equipment at NRC to study radiolysis of penicillamine disulfide and observed an analogous anion-radical. Further, by reducing the pH to 5.0 we were able to detect the spectrum of the thiyl free radical (RS.) and measure its rate of decay. However, in spite of all the gamma and pulse radiolysis work on these and related compounds nothing was found that could explain the large difference in radioprotective properties of various aminothiols.
In order to approach the problem from a different angle I spent a year at The Holt Radium Institute and Christie Hospital in Manchester, England, to train in radiation biology with Michael Ebert and Alan Tallentire at the University of Manchester. The Bacillus Megaterium spores being used there provided an excellent model for radiobiology studies. Survival curves, obtained by plotting the fraction of cells surviving against radiation dose, were used to demonstrate the effect of radiation on cell survival. Oxygen is critical in radiobiology as cells are about three times more sensitive to radiation when oxygen is present during radiation. On exposure to radiation the bacterial spores had survival curves that were linear and reproducible to several orders of magnitude in both aerated and deaerated suspensions. The aerated suspensions were about three times more sensitive than the deaerated suspensions. An interesting dose rate effect was discovered whereby anoxic spores were more sensitive at high dose rates. However, more remarkable was the response of the spores in the presence of the radioprotective compounds, cysteine and cysteamine. The slope of the survival curves were not affected in either aerated or deaerated conditions. By contrast, mammalian cells in aerated suspension are normally protected by these compounds with dose reduction factors of two or more. Thus the response of Bacillus Megaterium spores in the presence of aminothiols differed significantly from mammalian cells.
After returning to Ottawa I decided to use mammalian cells to continue my investigation of the mechanism of radioprotection with aminothiols. Dr. Otto Voss at T.N.O., Rijswijk, in The Netherlands provided me with human kidney (T-cells); a cell line that had been studied extensively there. A series of experiments confirmed that these cells responded in the expected fashion; the cells in a deaerated suspension were about three times less sensitive to radiation than the aerated cells.
WR 2721 is one of the more promising compounds synthesised in the program at the Walter Reed Army Institute of Research (WRAIR) aimed at finding an effective radioprotective drug suitable for use in humans. WR-2721 was later given the drug name "Amifostine". In order to investigate it under various conditions the free thiol, WR 1065, was also used as it is the actual protective agent in cells after the phosphate group covering the thiol has been removed by enzymes. These compounds have the following structures:
WR-2721: NH2CH2CH2CH2NHCH2CH2S-PO3H2WR-1065: NH2CH2CH2CH2NHCH2CH2SH
Various experiments carried out with mammalian cells using cysteamine and WR 1065 confirmed that the latter was more effective at protecting the cells. WR-1065 had a Dose Reduction Factor (DRF) of 2.9, superior to cysteamine DRF 2.3. However, more interesting experiments were carried out in cooperation with Henry Schneider at NRC using a microcalorimeter and an oxygen meter. The calorimeter showed a burst of heat when WR 1065 was added to a cell suspension. A further insight into the process was gained using an oxygen meter. Addition of WR-1065 to a suspension of T-cells in culture medium caused rapid removal of oxygen under various conditions. It was expected that the free thiol would react with oxygen to form disulfide but the oxygen depletion was more rapid than expected.
A further insight came at the First Conference on Radioprotectors and Anticarcinogens in 1982. Dr. Ziegler of the Clayton Foundation Biochemical Institute, Austin, Texas, presented a paper in which he described how glutathione reductase enzyme in cells rapidly converts disulfides such as the one from WR-1065 to free thiols. The thiol then immediately reacts with oxygen to reform a disulfide which in turn is reduced back to free thiols by the reductase enzyme. This cycle continues until all the oxygen is removed from the cell which makes it less sensitive to radiation. It will stay that way until the WR-1065 is eliminated.
That was a Eureka moment for me. I realized immediately that this would explain the process of radioprotection by aminothiols. It also explained the oxygen experiments I had performed and probably the lack of protection of Bacillus Megaterium spores. I discussed the subject with Dr. Ziegler to be certain I had understood his presentation correctly and corresponded with him later. It solved the long-standing problem of structure and protective ability involving such compounds. The research program at WRAIR had revealed that in order to be protective, an amino thiol has to have an unbranched chain of two or three carbon atoms between the sulfur and the nitrogen atoms. When the chain is extended to four or five carbons the protection is lost. With a chain length of six or seven there is slight protection, probably due to a conformation that places the nitrogen and sulfur atoms in close proximity. Thus, it appears that the protective aminothiols muat have structures that mimic the sulfur-containing section of glutathione in order to react with the enzyme.
Back in Ottawa I prepared a paper combining the results of the calorimetry and oxygen studies with the new knowledge obtained from Dr. Ziegler. The conclusion was clear to me: The principal mode of action of aminothiol protective agents was due to removal of oxygen. A small additional degree of protection was attributed to hydrogen donation. However, in view of more recent results it is likely that the additional protection above that due to anoxia comes from binding of the aminothiol to DNA.
The paper was submitted to the International Journal of Radiation Biology in 1982 with Henry Schneider as co-author along with our assistants Elizabeth Inhaber and John Labelle. I was astonished when the report came back from the journal. The referee stated that there was no relationship between structure and radioprotective activity; hydrogen donation explained everything! He demanded that two sentences in the paper on structure/activity be deleted. I considered trying to convince the editor that the referee was totally wrong and ignorant of the known facts but that would have delayed publication. Further, I was certain that scientists working in the field would be well aware of the structure-activity relationship and realize that the mechanism I described solved that long-standing mystery. To cut a long story short I deleted the offending sentences and the paper was published in Int. J. Radiation Biol., 1983, Vol. 43, No. 5, 517-527. It inspired many research projects and citations continued for over three decades. I am grateful to the many scientists who took up the challenge and confirmed my explanation of the mechanism of radioprotection by aminothiols with a variety of elegant studies.
Amifostion was later approved for use in oncology and is used in special cases to protect sensitive tissue during radiotherapy.