INT . J . RADIAT . BIOL .,

1975,

VOL .

28,

NO .

3, 225-241

Int J Radiat Biol Downloaded from informahealthcare.com by University of British Columbia on 10/29/14 For personal use only.

Radiotoxicity of intracellular 67Ga, 125 1 and 3 H Nuclear versus cytoplasmic radiation effects in murine L1210 leukaemia KURT G . HOFER, CLIFTON R . HARRIS and J . MARSHALL SMITHf Institute of Molecular Biophysics, Florida State University, Tallahassee, Florida 32306 (Received

19 June 1975 ;

accepted 21 July

1975)

L1210 leukaemia cells were labelled with various doses of 87 Ga-citrate, 3 H-thymidine, or 125 1-iododeoxyuridine to evaluate the cytocidal effects of the intracellular decay of the three radionuclides . Based on radioisotope incorporation data, cellular dimensions, and intracellular radioisotope distributions ( 3 H and 1251 intranuclear, 87Ga cytoplasmic) the rates of deposition of cellular, nuclear, and cytoplasmic energy were calculated . In terms of energy absorption/cell, 87 Ga (LD 5o : 2250 keV/hr ; 69 rad/hr) was much less toxic than either 3H (LD50 : 325 keV/hr ; 10 rad/hr) or 1251 (LD5o : 50 keV/hr ; 1 . 5 rad/hr) . In terms of energy absorption/nucleus, 87 Ga and 3H produced almost identical effects (LD50 : 230 versus 255 keV/hr ; 22 . 2 versus 246 rad/hr), but 1251 remained much more toxic (LD 50 : 40 keV/hr ; 3 . 9 rad/hr) . These findings indicate that, although decay by electron capture in the cell nucleus (1251) is highly destructive, the same type of decay occurring in the cytoplasm ( 87 Ga) is ineffective in killing L1210 cells . An analysis of the data suggests that the cytotoxic effects of the three radioisotopes result exclusively from nuclear damage . Cytoplasmic absorption of radiation energy appears to contribute little, if anything, to the lethal effects of ionizing radiations . 1.

Introduction It is generally believed, but not rigorously proven, that the cytotoxic effects of ionizing radiations on mammalian cells result from damage to the DNA rather than from cytoplasmic absorption of radiation energy . Little or no experimental evidence has been provided to support the hypothesis of Bacq and Alexander (1961) that cell-death may be caused by radiation effects on extranuclear structures such as lysosomes . Irradiation of the cytoplasm with microbeams generally is not very effective in killing mammalian cells (Munro 1970) . Further, the dose required to produce damage in enucleated red blood cells is several orders of magnitude higher than the doses known to cause cell-death after irradiation of the nucleus (Schiffer, Atkins, Chanana, Cronkite, Greenberg, Johnson, Robertson and Stryckmans 1966) . Additional support for the concept that the nucleus is the radiosensitive target comes from experiments with intracellular radio-isotopes . For example, tritiated thymidine ( 3 H-TdR), which is incorporated into the DNA of proliferating cells, is far more toxic to mammalian cells than is evenly-distributed t Present address : Radiobiology Division, College of Medicine, University of Utah, Salt Lake City, Utah 84132.



Int J Radiat Biol Downloaded from informahealthcare.com by University of British Columbia on 10/29/14 For personal use only.

226

K. G . Ho f er et al .

tritiated water (Painter, Drew and Hughes 1958) . In bacterial systems, the killing efficiency per disintegration of tritiated thymidine is nearly three times greater than tritium activity introduced via protein precursors, such as tritiated leucine or histidine (Person 1963, Rachmeler and Pardee 1963) . It therefore seems likely that, at least in the case of beta-emitting radio-isotopes, the genetic material of the cell nucleus constitutes the most radio-sensitive target within the cell. However, since the maximum range of 3H beta-particles in biological material is about 6 microns, the decay of 3H in the mammalian cell irradiates both the nucleus and the cytoplasm . Thus, the possibility of cytoplasmic contributions to the cellular radiation damage cannot be completely excluded . A better indication of nuclear versus extranuclear damage might be obtained with radionuclides which decay by electron capture . This type of decay frequently is accompanied by a burst of low-energy electrons (Auger electrons, sometimes also Coster-Kronig electrons) due to the ` vacancy cascade ' occurring in the perturbed parent atom (Gillespie, Orr and Greig 1970) . The emission of such a dense shower of low-energy electrons results in a highly-charged daughter atom and a high density of electron irradiation in the immediate vicinity of the disintegrating radionuclide . Thus, decay by electron capture should produce highly-localized damage, regardless of the cell structure in which the radioisotope is located . The most extreme example of such an electron-emitting radio-nuclide in common use today is iodine-125 . We have shown elsewhere that 1251, incorporated as the thymidine analogue iododeoxyuridine ( 126 1UDR) into the DNA of proliferating murine leukaemia cells, causes radiotoxic effects which exceed those of DNA associated tritium by a factor of 12 to 15 (Hofer, Prensky and Hughes 1969, Hofer, DiBenedetto and Hughes 1970, Hofer and Hughes 1971), even though the ratio of intranuclear energy deposition per decay is only 2 .7 (see table 3 of Appendix) . Other investigators have demonstrated similar toxic manifestations of DNA-associated 1251 in mammalian tissue culture cells (Feinendegen, Ertl and Bond 1971, Roots, Feinendegen and Bond 1971, Burki, Roots, Feinendegen and Bond 1973, Prince and Adelstein 1973), bacteria (Ahnstrom, Ehrenberg, Hussain and Natarjan 1972) and viruses (Krisch 1972) . Because of the extraordinary toxicity of 1251 we decided to investigate the biological toxicity of gallium-67, another commonly-used radionuclide which decays by electron capture . Unlike 125 1, which is exclusively associated with the DNA of dividing cells (Hughes, Commerford, Gitlin, Krueger, Schultze, Shah and Reilly 1964, Hofer and Hughes 1970), most of the intracellular 67Ga appears to be located in the cytoplasm (Swartzendruber, Nelson and Hayes 1971, Glickson, Ryel, Bordenca, Kim and Gams 1973) . As a result, the intracellular decay of 1251 should produce high local concentrations of radiation energy within the cell nucleus, whereas the decay of 67Ga should primarily irradiate the cytoplasm . Here we try to evaluate the effects of nuclear versus cytoplasmic irradiation, to calculate the deposition and microdistribution of radiation energy in 125 1UDR, 3H-TdR, and 67Ga-labelled L1210 cells, and to relate the intracellular dosimetry calculations to the experimentally-determined cytocidal effects of the three radionuclides .



Toxicity o f intracellular 67Ga, 1251, 3H 2.

227

Materials and methods

Int J Radiat Biol Downloaded from informahealthcare.com by University of British Columbia on 10/29/14 For personal use only.

L1210 mouse lymphoid leukaemia cells were maintained in the peritoneal cavity of 8-10-week-old C3D2JF1 (C3H x DBA/2J) female mice by weekly passage of 10 6 cells . Four days after inoculation the L1210 cells were labelled with 1 µCi 125 1UDR/mouse, administered intraperitoneally (i .p .) in four injections of 0 . 25 µCi/mouse . The four 125 1UDR injections were given at successive intervals of 3 hours ; i .e . the total labelling time was 9 hours . This is considerably longer than required to assure complete labelling of logarithmicallygrowing L1210 cells (Yankee, DeVita and Perry 1967, Hofer and Hofer 1971) . The 125 IUDR used in these studies (specific activity : 2 . 2 x 10 6 mCi/mM) was prepared in our laboratory according to a modified version of the procedure described by Hughes et al. (1964) . Four hours after completion of the 125 IUDR labelling course the mice were killed in a chloroform chamber, and the tumour cells were harvested under sterile conditions by repeated rinses of the peritoneal cavity with heparinized Earle's balanced salt solution . Cells from 8-10 mice were pooled and washed twice by gentle centrifugation at 500 r .p .m . for 10 min to remove 1251 activity not bound to the tumour cells . During these procedures the cell suspension was kept continuously at 4°C . The final cell-sample was essentially free of cell debris . When fractions of the final cell suspension were precipitated with cold 10 per cent trichloroacetic acid, less than 3 per cent of the radioiodide was found in the supernatant fluid, indicating that more than 97 per cent of the total 1251 activity was associated with the DNA of the tumour cells (Hofer and Hughes 1970) . After the last centrifugation the cell suspension was adjusted to a concentration 200 million L1210 cells per ml and 0 . 25 ml aliquots (50 million cells) from this suspension were inoculated i .p . into non-radioactive ' intermediary ' host mice . Eighteen hours after transfer of the 125 1UDR-labelled L1210 cells into intermediary hosts, the cells were labelled with 67 Ga . The 67 Ga labelling doses ranged from 0 to 2000 µCi/mouse, administered as a single intraperitoneal injection . Carrier-free E 7 Ga (specific activity : 4 x 10 7 mCi/mM) was obtained as 67 Ga-chloride in dilute HCl (pH 2 . 0) from the Isotopes Division of the Oak Ridge National Laboratories . Immediately before injection the 67 Ga-chloride solution was neutralized with appropriate amounts of NaOH, and sufficient amounts of sodium citrate were added to the solution to produce a citrate concentration of 7 mg/kg body-weight . Three hours after administration of 67 Ga-citrate, the intermediary hosts were killed and the 1251-67Ga double-labelled cells were harvested as described above . The tumour cells obtained from each mouse were kept in separate test-tubes and washed four times by gentle centrifugation to remove 67 Ga activity not incorporated into cells . When the radioactivity of the supernatant fluid had declined to insignificant levels, the tumour cells were resuspended in fresh Earle's balanced salt solution and 10 million double-labelled L1210 cells were inoculated intraperitoneally into groups of five non-radioactive C3D2JF1 mice (final test hosts) . One or two days before the final test mice were inoculated with the labelled tumour cells, their drinking-water was supplemented with 0 . 1 per cent sodium iodide to depress subsequent accumulation of 1251 in the thyroid . Immediately after tumour inoculation and at daily intervals thereafter the whole-body 1251



K . G. Ho fer et al .

228

Int J Radiat Biol Downloaded from informahealthcare.com by University of British Columbia on 10/29/14 For personal use only.

radioactivity of the test hosts was monitored by counting individual live mice in a well-type crystal scintillation counter (well diameter 45 mm) . Previous studies had shown that the rate of 1251 excretion from mice bearing 125 IUDRlabelled L1210 cells can be used as an index of natural and artificially-induced death and breakdown of tumour cells in vivo (Hofer et al . 1969, Hofer 1970, Hofer and Hofer 1971) . Thus, if the presence of radiotoxic quantitites of 67 Ga in 125 1-labelled cells should cause any increase in the rate of cell-death, this would manifest itself in the form of a corresponding increase in the rate of 12,11 excretion from mice bearing 1251-67Ga labelled cells (Hofer and Hughes

1971) . To determine the excretion time of labelled breakdown products after cell-death

in vivo, control groups of mice were inoculated intraperitioneally with 125 IUDR labelled L1210 cells which had been killed by prior in vitro irradiation with 5000 rad from a 137 Cs radiation source, or by heating in a 60°C water-bath for 12 min . As with host animals bearing live labelled L1210 cells, mice inoculated with dead labelled L1210 were counted each day in a well-type scintillator . All counts were corrected for radioactive decay and day-to-day fluctuations in the efficiency of the counting equipment by monitoring appropriate standard samples of 1251 and 67 Ga . In addition, in the case of 1251-67G a double-labelled cells, the wholebody 11,11 retention data had to be corrected for a 4. 6 per cent overlap of the 67 Ga spectrum into the 1251 channel . The somewhat involved double-labelling scheme outlined above was necessary to permit external monitoring of the death of tumour cells in living animals . Unlike 125 1, which is rapidly excreted after cell-death, large fractions of the 67 Ga activity released by cell-death are reutilized in various host tissues (Hofer and Swartzendruber 1973) . Consequently, the rate of 67 Ga excretion from tumour-bearing mice does not constitute a measure of the death of tumour cells in vivo, and trace quantities of 11,1 1 had to be present in 67 Ga-labelled cells to indicate the extent of 67 Ga-induced cell-lethality. The studies of cytotoxicity in vivo described above consumed only about one-half to two-thirds of the double-labelled cells harvested from the intermediary hosts . Samples of the remaining cell suspensions were retained to permit autoradiographic evaluation of the labelling index . Additional samples were reserved for determining the exact quantities of 1251 and 67 Ga present in double-labelled cells . To obtain data for microdosimetry calculations, we also carried out a series of cell-fractionation experiments designed to disrupt 1251-67Ga labelled cells and strip the cytoplasm from the nuclei by density centrifugation . The double-labelled cells were incubated for 15 min in 0 . 01 M tris buffer containing 1 per cent Tween 80 detergent . The cell suspension was then homogenized by hand with 30 strokes of a Teflon pestle in a 10 ml glass homogenizing chamber . The resulting mixture of whole cells, cell nuclei, and cytoplasmic fragments was washed twice in 0 . 01 M tris-detergentbuffer by centrifugation at 800 r .p .m . The final pellet was resuspended in 0 . 15 M sucrose and layered over a discontinuous sucrose gradient which consisted of three consecutive layers of sucrose (0. 8 M, 1 . 6 M, and 2 . 2 M) dissolved in 0 . 01 M tris/detergent-buffer. The discontinuous gradient was centrifuged at 3000 r .p .m . (about 1700 x g from the middle of the tube) for 5 min at 4 ° C . After centrifugation the fractions were collected with a Pasteur pipette . The 2 . 2 M sucrose layer routinely consisted of 95-100 per cent ' clean nuclei ' .



Int J Radiat Biol Downloaded from informahealthcare.com by University of British Columbia on 10/29/14 For personal use only.

Toxicity o f intracellular

87

Ga,

1251,

3H

229

To permit accurate dosimetry calculations, careful attention was devoted to the proper calibration of our radiation-counting equipment . The well-type crystal scintillation counter used in this investigation was calibrated with the help of radiation standards purchased from the New England Nuclear Corporation . The counting efficiency of our scintillation system was 54 per cent for 1251 and 16 per cent for 67 Ga . As mentioned above, trace amounts of 1251 were present in all 67 Ga-labelled samples . The presence of 1251 did not disturb the measurement of 87 Ga, because the low energy photons of 1251 did not overlap into the 67 Ga channel . However, there was a 4 . 6 per cent overlap of 67 Ga spectrum into the 1251 channel . With 67 Ga-labelling doses of 250 µCi/mouse or less, this overlap could be easily eliminated by counting the double-labelled cells (or mice bearing double-labelled cells) with two pre-set windows and subtracting 4 . 6 per cent of the 67 Ga counts from the 1251 counts . At higher 67 Ga-labelling doses, the overlap counts contributed a substantial fraction of the total count rate registered in the 1251 channel . To ascertain the accuracy of the overlap correction procedure at such high 67 Ga doses, the 67 Ga present in double-labelled cells was extracted by precipitation with 10 per cent cold trichloroacetic acid (Hofer and Swartzendruber 1973) . Groups of mice bearing L1210 cells labelled with more than 250 µCi of 67 Ga were killed at different stages of tumour development, followed by homogenization in a Virtis mixer and precipitation of the whole body with 10 per cent trichloroacetic acid . In all cases the resulting ' pure ' 1251 data were almost identical to the data obtained by overlap correction . We believe, therefore, that all of our radioactivity measurements are accurate to ± 10 per cent . 3 . Results 3 .1 . Incorporation of fi 7 Ga-citrate into L1210 cells Table 1 shows the number of peritoneal L1210 cells and the percentage of the 67 Ga activity present in these cells at the time of transfer to non-radioactive test hosts, 3 hours after intraperitoneal administration of 25, 50, 100, 250, 500, 1000, or 2000 µCi of 67 Ga/mouse . Fractional 67 Ga uptake remained constant over the entire dose range studied . The number of peritoneal L1210 cells also showed no significant variations . Autoradiographic evaluation of labelled samples indicated that the incorporated B 7 Ga was evenly distributed amoung the labelled cells (labelling index 100 per cent) . It was therefore

87 Ga-citrate dose (µCi/mouse) 25 50 100 250 500 1000 2000

Cell number (10 6 /mouse)

44 53 55 51 56 48 49

67 Ga incorporation (per cent of injected dose) 9 . 24 9 . 55 10 . 69 9 . 54 9 . 78 8 . 12 8 . 35

Table 1

67 Ga content (µCi/10 6 cells)

Disintegrations/ cell-hr

0 . 052 0 . 090 0 . 194 0 . 468 0 . 873 1 . 692 3 . 408

6.9 12 . 0 25 . 8 62 . 3 116 . 3 225 . 4 453 . 9



230

K . G . Ho f er et al .

possible to recalculate the 67 Ga retention data and express them in terms of µ Ci/10 6 labelled cells (table 1, column 4) or disintegration per single-labelled cell per hour (table 1, column 5) .

Int J Radiat Biol Downloaded from informahealthcare.com by University of British Columbia on 10/29/14 For personal use only.

3 .2 . Intracellular distribution of 67 Ga in labelled L1210 cells To evaluate the microdistribution of radiation energy within labelled cells, it was necessary to determine both the radio-isotope content per cell (table 1) and the intracellular distribution of the incorporated radionuclides (table 2) . Using the cell-fractionation technique described in § 2, we carried out a series of cell-fractionation experiments with double-labelled L1210 cells . The final fraction of ` clean ' nuclei recovered after cell disruption end centrifugation in a discontinuous sucrose gradient usually contained about 50-75 per cent of the total number of nuclei present in the initial cell suspension . Thus, the total amount of radioactivity present in the nuclei could be determined by measuring the radionuclide content of the clean nuclear fraction and correcting for the partial loss of nuclei sustained during homogenization and gradient separation . Table 2 presents the results of one such fractionation study on L1210 cells labelled with both 125IUDR and 67 Ga-citrate . As expected, more than 95 per cent of the incorporated 1251 was associated with the nuclei of the doublelabelled L1210 cells . In contrast, only 5 . 2 per cent of the intracellular 67 Ga

Fraction

1251 C .P .M .

L1210 cells

124,600

L1210 nuclei

118,500 (95 . 1 per cent)

B

7 Ga c .p .m . 52,300 2,700 (5 . 2 per cent)

Table 2 .

was recovered with the nuclear fraction . Since it is very difficult to obtain nuclei that are completely free of cytoplasm, it appears likely that the 5 . 2 per cent 67 Ga content of the nuclear fraction represents residual cytoplasmic contamination rather than actual retention of 67 Ga within the nuclei . Our findings thus confirm previous reports, which also indicated that 67 Ga administered as 67 Ga-citrate is retained largely in the cytoplasm (Swartzendruber et al . 1971, Glickson et al. 1973) . 3 .3 . Radiotoxic effects o f intracellular 67 Ga versus 1251 and 3H The experiment illustrated in figure 1 was designed to evaluate the viability of L1210 cells labelled with various amounts of 67 Ga, followed by inoculation into new, non-radioactive host mice . As pointed out in § 2, the 67 Ga-labelled cells were double-labelled with trace quantities of 125IUDR to permit external monitoring of the cytotoxic effects of incorporated 67 Ga . The minute quantities of 125 IUDR present in the double-labelled cells (0 . 004 j Ci/10 6 cells) were not sufficient to cause any cell-lethality (Hofer and Hughes 1971) . Thus, most



Toxicity

o

f

intracellular

87Ga,

125j, 3 H

23 1

or all of the cell-killing observed at high B 7Ga-labelling doses must have been due to the radiotoxic effects of the intracellular 87 Ga, even though the presence of the 1251 may have caused some additional sublethal effects . At low, non-toxic dose levels of B 7Ga the death rate among the transplanted L1210 cells showed the pattern observed in earlier studies on the same tumour line (Hofer al. 1969, Hofer and Hofer 1971) . About 10-15 per cent of the

Int J Radiat Biol Downloaded from informahealthcare.com by University of British Columbia on 10/29/14 For personal use only.

et

67

Ga dose (µCi/mouse) 60T t 0 .25 . Sa.ioo.z50 t 500

V O M

NN

1

2

1 I I I

4 6 Days after tumour inoculation

1

I

8

Figure 1 . Death of L1210 cells labelled with 1 µCi of 125 IUDR and 0-2000 µCi of 87 Gacitrate per mouse . Each data point represents the average 1251 retention value of five test mice, expressed as per cent of the inoculated 1281 activity . The variation within each treatment group was small (maximum ± 10 per cent of the mean) . The mean day of death is indicated by + .

inoculated 1251 activity was lost from the leukaemic mice each day during the initial phases of tumour growth . In advanced tumours, the rate of 1251 excretion, i .e . the rate of natural death of tumour cells in vivo, gradually declined to less than 10 per cent per day . Labelling doses of 500 µCi of B 7 Ga per mouse, or higher, progressively increased the rate of 1251 loss from recipient mice, usually after a lag phase of 1 or 2 days (figure 1) . On day 7, when most of the experimental animals died, low-dose groups still retained 55-60 per cent of the inoculated 1251 activity, whereas mice receiving L1210 cells labelled with 2000 µCi of 67 Ga-citrate retained only about 20 per cent of the initial 1251 activity .



K . G . Ho fer et al .

232

Figure 2 presents the 67 Ga toxicity data in the form of a conventional dosesurvival curve . Such curves can be calculated from whole-body 1251-retention data by interpolation between live and dead control values in the following manner :

T x -D(1-Tx /100) S

To - D(1 - To /100)'

Int J Radiat Biol Downloaded from informahealthcare.com by University of British Columbia on 10/29/14 For personal use only.

where S is the surviving fraction, T o is the fractional retention of 1251 in live control groups (labelled only with 125IUDR), T., means the 1251 retention of

1 .0

Figure 2 . Fraction of L1210 cells surviving various radio-isotope doses, plotted as a function of disintegrations /cell-hr. mice receiving L1210 cells labelled with various quantities of 67 Ga, and D stands for the whole-body retention of 1251 after inoculation of 131 Cs-killed 125 1UDRlabelled cells (D in this experiment was 1 . 9 per cent) . The experimental and mathematical basis for this equation were presented in detail in a previous report (Hofer and Hughes 1971) . Using this equation, the fractional survival of L1210 cells treated with various doses of 67 Ga was calculated and plotted as a function of disintegrations/cell-hr at the time of tumour transplantation (figure 2) . Similar radiotoxicity studies were previously carried out in our laboratory with tritiated thymidine ( 3 H-TdR), 131IUDR, and 125 IUDR (Hofer and Hughes 1971) . For comparison purposes, some of these results are included in figure 2 . From these data it was apparent 1251 is far more toxic to L1210 cells than either 3 H or 67 Ga . The number of that disintegrations/cell-hr required to reduce the surviving fraction to 50 per cent of the live control value is 4 for 125 1, 62 for 311, and 345 for 67 Ga . It should be pointed out, however, that distintegrations/cell-hr cannot be equated with dose-rate . The three radionuclides have very disparate decay energies, ranging from an average of 5 . 6 keV/decay for 3 H to 20 . 8 keV for 1251 and 32 . 1 keV for 67G a . 1251 and 67 Ga decay energies refer only to the combined



Toxicity

o f

intracellular 87Ga, 1251 , 3H

23 3

Int J Radiat Biol Downloaded from informahealthcare.com by University of British Columbia on 10/29/14 For personal use only.

energies of the electrons . The two radionuclides also emit electromagnetic radiations (X-rays and gamma-rays), but these types of radiation can be ignored in intracellular dosimetry calculations, because electromagnetic radiations generally have mean free paths which are at least an order of magnitude greater than cellular dimensions . Since the three radioisotopes also differ from each other with regard to their energy spectra and their intracellular distribution, the disintegrations/cell-hr values must be converted to rates of energy deposition per cell (or cell nucleus) to yield a meaningful measure of radioisotope toxicity (for details see Appendix) . Figure 3 illustrates the fractional survival of 67 Ga-labelled L1210 cells as a function of keV/cell-hr, rather than disintegrations/cell-hr . As can be seen by comparing figure 2 with figure 3, this procedure does not materially change the toxicity differential between the three radionuclides . In terms of energy deposition per cell, the dose required to cause 50 per cent cell-lethality s 50 keV/cell-hr (1 . 5 rad/hr) for 125 1, 325 keV/cell-hr (10 rad/hr) for 3H, and 2250 keV/cell-hr (69 rad/hr) for 87 Ga . The surprising lack of radiotoxicity from intracellular decay of 67 Ga prompted us to carry out further dosimetry calculations . As pointed out earlier in table 2, 67 Ga is located almost entirely in the cytoplasmic portion of the cell . Consequently, large fractions of the energy carried by the low-energy electrons emitted during 67 Ga decay are absorbed within the cytoplasm . A significant portion of the remaining radiation energy escapes from the labelled cells, and only a small fraction of the 67 Ga decay energy reaches the cell nucleus (see table 4 in Appendix) . Thus, if it is true that the cytocidal effects of ionizing radiations

Dose to the entire cell (rod/hr) 20 40 60 s0

125 I

3H

1 1 ) 1 1 1000 2000 3000

1

Dose to the entire cell (KeV/cell- hr)

Figure 3 .

Fraction of L1210 cells surviving various radio-isotope doses, plotted as a function of radiation energy/cell-hr (keV/hr or rad/hr) .



23 4

K . G . Hof er

et al .

Int J Radiat Biol Downloaded from informahealthcare.com by University of British Columbia on 10/29/14 For personal use only.

on mammalian cells result primarily from damage to the DNA (Marin and Bender 1963 a, b), the toxicity of 67 Ga should be proportional to the fraction of radiation energy which is deposited in the cell nucleus . To test this hypothesis, the cellular dose-survival values for the three radionuclides were recalculated, and the fractional survival was plotted as a function of dose-rate per nucleus rather than dose-rate to the whole cell (figure 4) . Using this procedure, the dose-survival curves for 87Ga and 3H become virtually superimposed, but the curve for 1251 remains displaced to the left . In terms Dose to the cell nucleus(rad/hr) .0

10

20

30

40

U. 4

N

1 I I 0 100 200 300 400 Dose to the cell nucleus(KeV/nucleus-hr)

Figure 4. Fraction of L1210 cells surviving various radio-isotope doses, plotted as a function of radiation energy/nucleus-hr (keV/hr or rad/hr) . of energy deposition per cell nucleus, the dose-rate required to produce 50 per cent cell-death is 40 keV/nucleus-hr (3 .9 rad/hr) for 125 1, 255 keV/nucleus-hr (24 .6 rad/hr) for 3H, and 230 keV/nucleus-hr (22 .2 rad/hr) for 87 Ga . These results indicate that in terms of energy deposition/nucleus 67Ga and 3H produce almost identical cytocidal effects . However, even after correcting for differential energy absorption in the nucleus, 1251 remains much more toxic than either 3H or 67Ga . 4. Discussion In previous communications we have demonstrated that the rate of 1251 (or 131 1) loss from mice bearing 125 1UDR (or 131 1UDR)-labelled L1210 cells provides a quantitative measure of spontaneous and therapy-induced death of tumour cells in vivo (Hofer et al ., Hughes 1969, Hofer 1970, Hofer and Hofer 1971) . The IUDR pre-labelling technique has also proved to be a sensitive assay for evaluating the cytocidal effects resulting from intracellular decay of incorporated radionuclides (Hofer and Hughes 1971) .



Toxicity of intracellular s'Ga, 125J, 3H

235

Ideally, radio-isotope-induced cell-death should be evaluated as a function

Int J Radiat Biol Downloaded from informahealthcare.com by University of British Columbia on 10/29/14 For personal use only.

of cumulative dose per cell or cell nucleus . Unfortunately, insufficient knowledge concerning the growth kinetics of individual labelled L1210 cells in vivo makes it very difficult to calculate cumulative dose values . Since the tumour cells are labelled by a single radio-isotope injection, the total radiation dose accumulated per cell (or cell nucleus) during the period between labelling and first cell division depends largely on the position of the cell in the cell-cycle at the time of labelling . In other words, cells labelled during the early phases of the cycle will be exposed to the high initial radiation dose-rate for the full length of the cell-cycle (9 hours for logarithmically-growing L1210 cells), whereas cells labelled at the end of the cycle experience such high dose-rates for only short periods before dividing and distributing their label among the two daughter cells . Consequently, the radiation dose accumulated during the first cell-cycle will not be constant, even if all labelled cells contain equal amounts of radioactivity . Of course, the radiation load accumulated during subsequent generation cycles will be the same for all labelled cells, regardless of the time of radio-isotope incorporation . Variations in the length of the cell-cycle induced by the experimental procedures introduce additional complications . For instance, L1210 cells transplanted into a new host experience a 9-12 hour ` lag phase ' before resuming cell proliferation (Hofer and Hofer 1971) . In addition, heavily-labelled tumour cells may experience an appreciable degree of further mitotic delay due to irradiation from the incorporated radioisotope . Previous data (Painter and Drew 1959, Drew and Painter 1962, reviewed in Bond and Feinendegen 1966) indicate that tritium doses sufficient to cause 99 per cent lethality among 3H-TdRlabelled HeLa cells produce division delays of about 7 hours . Assuming that radio-isotope effects in vivo are similar to those observed in vitro, the lag phase of heavily-labelled L1210 cells may be prolonged by several hours . Both these effects would tend to potentiate the radiation damage, since the dilution of the label would be retarded and the total number of disintegrations accumulated per cell would be correspondingly enhanced . The experiments reported in this communication were designed to minimize the importance of these factors by monitoring the radiation effects of different radio-isotopes under identical experimental conditions . In other words, although we did not evaluate the length of the radiation-induced division delay or the duration of the lag phase after transplantation into new hosts, it can be assumed that the growth pattern was similar in all corresponding experimental groups . Thus, the rate of energy deposition in the cell or cell nucleus at the time of tumour transfer does permit valid comparisons on the relative biological effectiveness (r .b .e .) of the three radio-isotopes . When the surviving fraction of various radio-isotope labelling groups is plotted as a function of the rate of energy deposition in entire cells, the resulting dose-survival curves indicate a vast toxicity differential between the three radionuclides . In terms of dose-rate to the entire cell 1251 is much more toxic than 3H, which in turn is more toxic than 6'Ga (figure 3) . However, when the cytoplasmic energy deposition is disregarded and fractional survival is expressed as a function of nuclear energy deposition, it becomes apparent that the toxic effects of 3 H and 6'Ga are virtually identical (figure 4) . 125 1, on the other hand, remains much more radiotoxic than either of the other two radionuclides .

K . G. Ho f er et al.

236

Int J Radiat Biol Downloaded from informahealthcare.com by University of British Columbia on 10/29/14 For personal use only.

Two different hypotheses have been advanced to explain the excessive radiotoxicity of intranuclear 125 1. Both are based on the fact that the electron capture decay of 1251 is accompanied by a dense shower of Auger electrons . As shown in table 3, ten electrons are emitted on the average during the decay of 125 1 . Seven of the ten electrons have an initial energy of 0 .4 keV or less . Such low-energy electrons have a maximum range of about 200 A, which corresponds to a nominal LET of at least 20 keV/µ . However, the effective LET or `linear

keV 125I

keV

(Gillespie, Orr and Greig 1970)

87 Ga (Dillman 1970)

Energy (keV)

Ave . number per decay

Energy (keV)

34. 5 30 . 9 30 . 5 29 . 0 26 . 3 22 .7 3. 7 3. 0 0. 4 0. 3 0. 2

0.040 0.010 0.100 0.023 0.060 0. 139 0 . 780 1.654 3 . 550 3 . 550 0. 290

378 . 3 286 . 3 196 . 3 183 .4 174 . 8 93 . 2 92 . 2 90 .1 83 . 6 81 .5 9. 5 8. 5 7. 5 0. 9 0. 1

Ave . number per decay 0 .00015 0.0006 0.0004 0.00015 0.0025 0. 0089 0. 0267 0. 00015 0. 2496 0. 0017 0. 0067 0. 1397 0. 5136 1 . 73 3. 68

Table 3 . energy ', which is a more useful concept for low-energy electrons, is probably several times higher . This quantity is defined as the energy deposited in a region divided by the mean chord length of the region (ICRU 1971) . Lowenergy electrons undergo an extreme degree of scatter, and their effective LET may therefore easily approach 100 keV/µ (Cole 1969) . In addition, since the Auger electrons in a cascade are emitted within 10 -15 sec (Wexler 1967) and the half-life of the 125 T,- daughter is 1 . 6 nanosec (the decay of which gives rise to a second, independent Auger shower), the cumulative local energy density at the site of 1251 decay may be equivalent to several hundred keV/µ . In other words, the specific ionization density at the site of 1251 decay within the DNA double helix is probably equivalent to the passage of a particle with very high LET (100 keV/µ or more) through that region . It is not unreasonable to assume that the r .b .e. of such an event must be very high . According to a second hypothesis, the biological toxicity of intranuclear 1251 may be caused by molecular fragmentation rather than by deposition of radiation energy. The electron capture decay of 1251 and the resulting vacancy cascade cause the appearance of a tellurium daughter with 1 to 18 (average 9) positive charges (Carlson and White 1963) . This high concentration of positive



Toxicity o f intracellular 6'Ga,

125 1 ,

3H

237

Int J Radiat Biol Downloaded from informahealthcare.com by University of British Columbia on 10/29/14 For personal use only.

charges is dispersed throughout the molecule by intramolecular electronic attraction, and the resulting unstable distribution of positive charges may cause molecular fragmentation by a ` coulombic explosion' . This process has been confirmed in the gas phase for small organic molecules labelled with 1251 (Carlson and White 1963, Wexler 1967), but it is not yet clear whether this mechanism operates in a condensed phase or within macromolecules . Support for either hypothesis can be derived from the findings of Krisch and Ley (1973), who demonstrated that 1251 incorporated into the DNA of bacteriophage T4 produces double-strand breaks with almost 100 per cent efficiency. These data suggest that phenomena local to the site of decay are responsible for the radiotoxicity of 125 1 . Unfortunately, evaluation of the number of double-strand breaks per decay does not explain the mechanism of 125 1 toxicity. It therefore cannot be ruled out that the high LET-type action of DNA-bound 1251 may be due to a combination of factors, such as high local ionization density, molecular fragmentation, and chemical transmutation . Whatever the mechanism of 1251 toxicity may be, our data, as well as the findings of other investigators, indicate that radio-isotope decay by electron capture in the cell nucleus is extremely toxic to mammalian cells . In contrast, cytoplasmically located 67 Ga, in spite of the similarity in the decay pattern, is surprisingly non-toxic . It can therefore be concluded that even highlyconcentrated deposition of radiation energy in the cytoplasm is not very effective in killing mammalian cells . This finding has interesting biological implications . For example, if the microdistribution of radiation energy in 67 Ga-labelled cells is compared with that of gamma- or X-irradiated cells, it becomes apparent that cytoplasmic absorption of radiation energy cannot be a major contributor to cell injury from external irradiation . From the data shown in the Appendix, it can be calculated that in 67 Ga-labelled L1210 cells the cytoplasm is irradiated at almost nine times the rate of the cell nucleus ; i .e . 8 . 77 keV of radiation energy are absorbed in the cytoplasm for each 1 keV of nuclear energy deposition . With evenly-distributed external radiations, the cytoplasm of L1210 cells receives only 2 . 0 keV for each 1 keV of nuclear energy deposition . It is therefore obvious that L1210 cells irradiated with X-rays or gamma-rays will receive a lethal nuclear dose long before they come even close to accumulating a damaging dose level in the cytoplasm . By implication, these calculations also disprove the ' enzyme release hypothesis ' of Bacq and Alexander (1961) . This hypothesis suggests that radiation may kill cells by damaging lysosomal membranes (or other membranous barriers), which in turn might cause cell lysis due to release of proteolytic and/or nucleolytic enzymes . Because of the low toxicity of cytoplasmic 67 Ga, it seems extremely unlikely that lysosomes or any other extra-nuclear structure could be the primary target for radiation-induced cell-lethality . The low radiotoxicity of intracellular 67 Ga also has certain implications in the area of nuclear medicine . Radio-isotopes are frequently mentioned as possible curative agents for cancer, particularly in cases where a radioisotope is known to accumulate within or in the vicinity of cancer cells (Greig, McDougall and Halnan 1973) . Although 67 Ga, administered as 67 Ga-citrate, has been found to concentrate in a variety of human and animal tumours (Hayes, Nelson, Swartzendruber, Carlton and Byrd 1970, Swartzendruber, Byrd, Hayes, Nelson R .B .

S



238

K . G . Ho fer et al .

and Tyndall 1970), our findings indicate that the radiotoxicity of incorporated 67 Ga is so surprisingly low that this radionuclide has little potential as a radiocurative agent in cancer .

Int J Radiat Biol Downloaded from informahealthcare.com by University of British Columbia on 10/29/14 For personal use only.

5 . Conclusion In terms of radiation energy deposited per cell intranuclear 1251 and 3 H are much more damaging to L1210 leukaemia cells than cytoplasmically-located 67 Ga. However, when the cytocidal effects of the three radionuclides are evaluated as a function of energy deposition in the cell nucleus, the dose-survival curves of 3H and 67 Ga become virtually superimposed . In other words, although 67 Ga is located primarily in the cytoplasm, its radiotoxic effects can be explained on the basis of overlap radiation energy deposited within the cell nucleus . Most or all of the radiation energy absorbed in the cytoplasm appears to be ' wasted ' . We can therefore conclude that, at least in mammalian cells, the cytoplasmic contribution to cellular radiation injury, if any, must be minimal . Appendix The ranges of many of the electrons in the electron spectra of the radioisotopes considered in this work are on the same order as the dimensions of the cell . Intracellular dosimetry calculations therefore must be based on careful evaluation of the cellular geometry, the intracellular distribution of the radioisotopes, and the details of the propagation of the emitted electrons . According to Berger (1970) the fraction 0 of energy emitted by point isotropic emitters uniformly distributed in a source region which is absorbed in a given target region is given by 00 (1) O(Eo) = pVJP(x)o(x, E0) dx, 0

where O(x, E0) is the point isotropic specific absorbed fraction (fraction of energy emitted by the source which is absorbed per unit mass at a distance x from the source), V is the volume of the target region and p is its density, E 0 is the initial energy of the emitted particle, and P(x) is the pair-distance distribution function . This latter quantity is the probability distribution of the distances between pairs of points selected successively at random between the source and target regions (which may correspond) . To determine the P(x) values, a Monte Carlo FORTRAN programme was written . The average geometry of the L1210 cell was obtained by electron and optical microscopic observation . The cellular geometry chosen for our dosimetry calculations is represented by the cross-section shown in figure 5 rotated about the z axis . The cell is assumed to be a sphere of 5 . 00 µ radius, and the shape of the nucleus is approximated by a sphere of radius 3 . 62 µ concentric with the cell and with a paraboloid of revolution cut out of the nucleus . The apex of the paraboloid is at the nuclear centre and its equation is

r sine 0 = (1 . 08 µ) cos 6.

(2)

With these parameters the subcellular regions were defined for the Monte Carlo programme and the appropriate P(x) distributions were generated using up to several million random numbers for a given source-target and configuration . The functions 0 (x, E0) in equation (1) above were calculated from the scaled point kernels tabulated by Berger (1973) for electrons in aqueous media . These



Int J Radiat Biol Downloaded from informahealthcare.com by University of British Columbia on 10/29/14 For personal use only.

Toxicity o f intracellular 17 Ga, 125 1, 3 H

239

L1210 CELL Figure 5 .

Cross-section of an average L1210 cell rotated about the z axis .

point kernels were determined by extensive Monte Carlo calculations, taking into account the variation in LET along the electron's path, scattering, energyloss straggling, and electron energies down to 0 . 5 keV . A separate FORTRAN programme was written for performing numerical integrations of equation (1), using Berger's tables and for integration over the electron energy spectrum of the given radioisotope .

Radio-isotope

Source distribution

Nuclear dose keV/decay (rad/ decay)

Cellular dose keV/decay (rad/ decay)

3 H Robertson and Hughes 1959)

Nucleus

3 . 98 (0 . 384)

5 . 17 (0 . 158)

1251

Nucleus

(Gillespie

10-6(1-03)

12 . 5 (0 . 384)

et al . 1970) 87 Ga (Dillman 1970)

Cytoplasm

0 . 666 (0 0644)

6 . 51 (0 . 200)

Table 4 . With the computer programme described above, the energy deposition and dose per decay to the entire cell and the cell nucleus were calculated for uniform distribution of 3H and 1251 in the nucleus and for a uniform distribution of 67Ga in the cytoplasm . The results are summarized in table 4 . To check the accuracy of our computer technique, we compared our results with those of Berger (1973), which were obtained using an analytically derived pair-distance distribution for 3 H and 1251 self-absorbed in spheres of varying diameter, and found agreement to within 5 per cent .



240

K. G . Hofer et al . ACKNOWLEDGMENTS

Int J Radiat Biol Downloaded from informahealthcare.com by University of British Columbia on 10/29/14 For personal use only.

We are indebted to Mr . Raymond L . Warters for his assistance in carrying out cell-fractionation studies . Appreciation for technical assistance on this research is extended to Miss Abigail Rayner, Mrs . Maria Hofer, and Mr. Ralph Hamilton. This investigation was supported by NIH grant No . CA 13222-03 and by a contract between the Division of Environmental and Biomedical Science, U . S . Atomic Energy Commission, and the Florida State University. Les cellules L1210 de leucemie etaient etiquetees avec des doses diverses de 87Gacitrate, 3H-thymidine ou 125 1-iododeoxy-uridine pour evaluer les effets cytocides qui proviennent de la desintegration intracellulaire des trois radionuclides . On a calcule d'apres les donnees de l'incorporation radio-isotope, les dimensions cellulaires et les distributions radio-isotopes intracellulaires ( 3 H et 1251 intranucleaire, B 7 Ga cytoplasmique), les frequences de 1'energie de depot cellulaire, nucleaire et cytoplasmique, selon les proprietes connues de chaque isotope . En ce qui concerne 1'absorption d'energie par cellule, 67 Ga (LD 50 : 2250 keV/cell/hr) 69 rad/hr) etait beaucoup moins toxique que 3H (LD 60 : 325 keV/cell-hr ; 10 rad/hr) ou 1251 (LD 50 : 50 keV/cell-hr ; 1,5 rad/hr) . Par rapport a l'absorption d'energie par noyau, 67 Ga et 3H ont produit des effets presque identiques (LD 50 : 230 versus 255 keV/noyau-hr ; 22,2 versus 24,6 rad/hr), mais 1251 est reste beaucoup plus toxique (LD 50 : 40 kev/noyau-hr ; 3,9 rad/hr) . Ces decouvertes indiquent que, quoique la desintegration par la prise d'electrons dans le noyau cellulaire (1251) soit tres destructrice, la meme sorte de desintegration qui se produit dans le cytoplasme ( 67 Ga), est plutot inefficace pour tuer les cellules L1210. Une analyse des donnees suggere que les effets cytotoxiques des trois radio-isotopes proviennent exclusivement du dommage nucleaire . L'absorption cytoplasmique de la radiation parait peu contribuer aux effets mortels des radiations ionisantes . L1210 Leukamiezellen wurden mit unterschiedlichen Dosen von 67 Ga-Zitrat, 3 H125 J-Joddesoxyuridin markiert, um die durch den intrazellularen Zerfall Thymidin oder der drei Radioisotope hervorgerufenen Zellschaden zu ermitteln . Auf Grund von Untersuchungen fiber den Radioisotopeneinbau, die Zelldimensionen and die intrazellulare Verteilung der Radioisotope ( 3H and 125J im Zellkern, B7 Ga im Cytoplasma) konnte die Energieabgabe in der Gesamtzelle, im Zellkern and im Cytoplasma berechnet werden . Wird die Energieabsorption in der Gesamtzelle in Betracht gezogen, dann erscheint 67 Ga (LD 50 : 2250 keV/Std ; 69 rad/Std) wesentlich weniger schadigend als 3 H (LD50 : 325 keV/Std ; 10 rad/Std) oder 125 J (LD 50 : 50 keV/Std ; 1-5 rad/Std) . BerUcksichtigt man jedoch ausschlieBlich die Energieabsorption im Zellkern, dann verursach°n 67 Ga and 3H nahezu identische Strahlenschaden (LD 50 : 230 bzw. 255 keV/Std ; 22,2 bzw . 24,6 rad/Std), wahrend 125 J auch in dieser Hinsicht wesentlich destructiver ist (LD5o : 40 keV/Std ; 3,9 rad/Std) . Obschon sich also radioaktiver Zerfall durch Elektroneneinfang im Zellkern (125J) auBerst schadigend auswirkt, scheint dasselbe Zerfallsereignis im Cytoplasma (B7 Ga) relativ harmlos fur L1210 Zellen zu sein . Eine Analyse der Ergebnisse deutet darauf hin, daB die von den drei Radioisotopen hervorgerufenen Letaleffekte ausschliel3lich auf Zellkernschaden zuri ckzufiihren sind . Cytoplasmatische Absorption von Strahlungsenergie scheint nur wenig oder gar nichts zum Strahlungstod der Zelle beizutragen .

REFERENCES AHNSTRO :VI, G ., EHRENBERG, L ., HUSSAIN, S ., and NATARJAN, A . T ., 1972, Mutation Res ., 10,247 . BACQ, Z . M ., and ALEXANDER, P ., 1961, Fundamentals of Radiobiology, 2nd edition (New York : Pergamon Press), Chaps . 9 and 10 .



Toxicity o f intracellular s'Ga, 125 1, 3 H

241

BERGER, M . J ., 1970, Medical Radionuclides : Radiation Dose and Effects, (U .S .A .E .C . Report CONF-691212), p . 63 ; 1973, National Bureau of Standards Internal Report,

No . 73-107 . BOND, V . P ., and FEINENDEGEN, L. E ., 1966, Hlth . Phys ., 12, 1007 . BURKI, H . J ., ROOTS, R., FEINENDEGEN, L . E ., and BOND, V . P ., 1973, Int . 5' . Radiat . Biol .,

24,363 .

Int J Radiat Biol Downloaded from informahealthcare.com by University of British Columbia on 10/29/14 For personal use only.

CARLSON, R. R ., and WHITE, R . M ., 1963, 5'. chem . Phys ., 38, 2930 . COLE, A ., 1969, Radiat . Res., 38, 7 . DILLMAN, L . T ., 1970,5' . nucl. Med ., 11, Suppl . 4, 17 . DREW, R . M ., and PAINTER, R . B ., 1962, Radiat. Res ., 16, 303 . FEINENDEGEN, L. E ., ERTL, H . H ., and BOND, V. P ., 1971, Proceedings of the Symposium on Biophysical Aspects of Radiation Quality (Vienna : IAEA), p . 419 . GILLESPIE, F . C ., ORR, J . S ., and GREIG, W . R., 1970, Br . Y. Radiol ., 43, 40 . GLICKSON, J . D ., RYEL, R . B ., BORDENCA, M . M ., KIM, K . H ., and GAMS, R. A., 1973,

Cancer Res ., 33, 2706 . GREIG, W . R ., MCDOUGALL, I . R ., and HALNAN, K . E ., 1973, Br . Med. Bull ., 27, 63 . HAYES, R . L., NELSON, B ., SWARTZENDRUBER, D . C., CARLTON, J . E ., and BYRD, B .

1970, Science, N.Y., 167, 289. HOFER, K . G ., PRENSKY, W., and HUGHES, HOFER, K . G ., 1970, Radiat . Res ., 43, 663 .

L .,

W . L ., 1969, 5'. natn . Cancer Inst., 43, 763 .

K . G ., DIBENEDETTO, J ., and HUGHES, W . L ., 1970, Z. Krebsforschung, 75, 34 . K . G ., and HOFER, M ., 1971, Cancer Res., 31, 402 . HOFER, K . G ., and HUGHES, W . L ., 1970, Cancer Res ., 30, 236 ; 1971, Radiat . Res., 47, 94 . HOFER, K . G ., and SWARTZENDRUBER, D . C ., 1973,5'. natn . Cancer Inst ., 50, 1039 . HUGHES, W . L ., COMMERFORD, S . L ., GITLIN, D ., KRUEGER, R . C ., SCHUTZE, B ., SHAH, V., HOFER, HOFER,

and REILLY, P ., 1964, Fed. Proc ., 23, 640 . INTERNATIONAL COMMISSION ON RADIATION UNITS AND MEASUREMENTS,

1971,

Report No

19,

Radiation Quantities and Units . KRISCH, R . E ., 1972, Int . S' . Radiat . Biol., 21, 167 . KRISCH, R. E ., and LEY, R . D ., 1973, 7th International Hot Atom Chemistry Symposium,

Julich, Germany . MARIN, G ., and BENDER, M . A ., 1963 a, Int . 5'. Radiat . Biol ., 7, 221 ; 1963 b, Ibid ., 7, 235 . MUNRO, T. T., 1970, Radiat . Res ., 42,451 . PAINTER, R . B ., andIDREw, R . M ., 1959, Lab . Invest ., 8, 278 . PAINTER, R . B ., DREW, R . M ., and HUGHES, W . L ., 1958, Science, N. Y., 127, 1244 . PERSON, S ., 1963, Biophys . 5'., 3, 183 . PRINCE, E . W ., and ADELSTEIN, S . J ., 1973, Radiat. Res ., 55, 543 . RACHMELER, M ., and PARDEE, A . B ., 1963, Biochim. biophys . Acta, 68, 62 . ROBERTSON, S . J ., and HUGHES, W . L ., 1959, Proc . of the First National Biophysics Conference (New Haven : Yale University Press) . ROOTS, R ., FEINENDEGEN, L . E ., and BOND, V. P ., 1971, Proceedings of the Third Symposium on Microdosimetry (Luxembourg : Euratom) . SCHIFFER, L . M ., ATKINS, H . L ., CHANANA, A . D ., CRONKITE, E . P ., GREENBERG, M . L ., JOHNSON, H . A ., ROBERTSON, SWARTZENDRUBER, D . C ., BYRD, B .

J . S ., and STRYCKMANS, P . A ., 1966, Blood, 27, 831 . L., HAYES, R . L ., NELSON, B ., and TYNDALL, R . L .,

1970, 5' . natn . Cancer Inst ., 44, 695 . SWARTZENDRUBER, D . C ., NELSON, B ., and HAYES, R . L., 1971,5'. natn . Cancer Inst., 46, 941 . WEXLER, S ., 1967, Science, N .Y., 156, 901 . YANKEE, R . A ., DE VITA, V . T ., and PERRY, S ., 1967, Cancer Res., 27, 2381 .