INT. J. RADIAT. BIOL., 1984, VOL. 45, NO. 5, 427-437

Constants of the Alper and Howard-Flanders oxygenequation for damage to bacterial membrane, deducedfrom observations on the radiation-inducedpenicillin-sensitive lesion

F. . OBIOHAt and N. E. GILLIES

Department of Anatomy and Biology as Applied to Medicine,The Middlesex Hospital Medical School, London WIP 6DB, U.K.

BEULAH M. CULLEN: and HILARY C. WALKER

M.R.C. Cyclotron Unit, Hammersmith Hospital, London W12 OHS, U.K.

TIKVAH ALPER

Birkholt, Sarisbury Green, Hampshire S03 6AL, U.K.

(Received 24 March 1983; revision received 20 December 1983;accepted 16 January 1984)

Energy deposited in the bacterial envelope of E. coli B/r induces lesions which arelethally attacked by penicillin in concentration insufficient to affect unirradiatedbacteria. The critical lesions are probably in the membrane moiety. Bacteria wereirradiated in the presence of 100 per cent oxygen, oxygen-free nitrogen andmixtures of 101, 0-59, 03, 0'1 and 0-06 per cent oxygen in nitrogen. Changes insensitivity with pO2 conformed with the Alper and Howard-Flanders equation,for bacteria treated after irradiation by penicillin as well as for the untreated ones.The values of m were respectively 4-8 and 33; the values of K were identical,within experimental error, i.e. 4-4 mmHg.

Sensitivity to induction of the penicillin-sensitive lesion was calculated fromthe difference in the reciprocals of Do values proper to untreated and treatedbacteria, for every gas used. The value of mn could not be directly calculatedbecause the effect of penicillin on anoxically irradiated bacteria was notdetectable. For that reason, a transformation of the oxygen equation was usedwhich allowed estimates to be made of both m and K, provided the resultsconformed with the equation. Within experimental error they did so conform.The calculated values of m and K for induction of the penicillin-sensitive lesionwere respectively 8 and 5-9 mmHg, but it is shown that the oxygen enhancementratio was probably underestimated and the K value overestimated.

On the assumptions that these values of m and K are specific for radiationdamage to bacterial membrane, and that radiation-induced killing is attributableto lethal lesions in the membrane as well as the DNA, the results demonstrate thatany interaction of oxygen with sites of energy deposition in the DNA must play avery much smaller role in radiosensitization than does interaction with sites ofenergy deposition in the membrane.

Indexing terms: bacteria, membranes, penicillin, oxygen effects, K values.

t Present address: Department of Radiation Medicine, University of Nigeria TeachingHospital, Enugu, Anambra State, Nigeria.

: To whom reprint requests should be addressed.

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Fr

eie

Uni

vers

itaet

Ber

lin o

n 11

/13/

14Fo

r pe

rson

al u

se o

nly.

F. I. Obioha et al.

1. IntroductionGillies et al. (1979) showed that, after irradiation of oxygenated E. coli B/r,

additional killing occurred when the bacteria were incubated with a concentration ofpenicillin that on its own did not kill the cells. No effect of penicillin was observedwhen the bacteria were anoxic during irradiation. Subsequently Gillies and Obioha(1982) demonstrated that penicillin caused the same effect in E. coli B/r, irradiatedunder anoxia, if certain highly lipophilic electron-affinic radiation sensitizers werepresent. The site of the antibiotic action of penicillin is known to be in thepeptidoglycan layer of the bacterial envelope (Blumberg and Strominger 1974)which is sandwiched between and attached to the outer and inner membranes of thecell envelope of E. coli (Braun and Sieglin 1970, Costerton et al. 1974). Because theantibiotic effect occurs in close proximity to the cell membrane and the efficiency ofradiosensitizers in combination with penicillin depends on their lipophilicity (Gilliesand Obioha 1982), it seems very probable that the radiation-induced lesion exposedby penicillin is associated with energy deposition within the membrane. Thisconclusion is reinforced by results of Obioha (1981) who demonstrated a similarlesion in bacteria treated after aerobic irradiation with antibiotics that act specificallyon the bacterial cell membrane.

The relationship between the partial pressure of oxygen (O 2) present in cells atthe time of X-irradiation and the extent of enhancement of the radiation effectinvolves two constants: m, the oxygen enhancement ratio (o.e.r.) and K, a measure ofthe rate of increase in enhancement with increasing pO2 (Alper and Howard-Flanders 1956). K is best expressed as a partial pressure of oxygen (Alper 1979).Knowledge of these constants can be useful in leading to some understanding of theunderlying mechanisms involved in the 'oxygen effect'. In this paper estimates aregiven of the values of m and K for specific damage to the membrane of E. coli B/r.

2. Materials and methods2.1. Bacteria

The strain used was E. coli B/r, which was maintained in the laboratory on slopesof Oxoid Blood Agar Base (BAB) kept at 4°C. For experiments, cells were inoculatedinto 10 ml of Oxoid nutrient broth and grown to stationary phase by incubation at37°C for 24 hours. The cells were harvested by centrifugation and resuspended in0-067 M phosphate buffer (pH 72) to give a concentration of 5 x 1' cells/ml.

2.2. Procedure for irradiationThe bacteria were exposed to electrons while lying on Millipore filters (0-22 tam;

12-5mm radius) in a gas-tight chamber (Alper et al. 1967a) through which theappropriate gas could be passed. The procedure was to filter ml of cell suspensionthrough a Millipore filter. Each filter could accommodate 5 x 107 cells in a singlelayer. Twelve filters were then placed on a thin layer of water-agar set in analuminium base which acted as the lid of the radiation chamber. The chamber wassealed tightly and the appropriate gas passed through it via inlet and outlet valves.The bacteria were irradiated through the 4 mm aluminium base of the chamber withelectrons from the linear accelerator in the Medical Research Council CyclotronUnit at Hammersmith Hospital. The energy of the beam was about 5 MeV at theposition of the bacteria and the field was uniform over the area of the chamber. Thetemperature in the chamber was about 22°C.

428

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Fr

eie

Uni

vers

itaet

Ber

lin o

n 11

/13/

14Fo

r pe

rson

al u

se o

nly.

Alper and Howard-Flanders constants

2.3. Gas conditions at the time of irradiationPure oxygen, 'white spot' nitrogen and the required mixtures of the two were

bought from The British Oxygen Company. With the exception of 100 per centoxygen, the concentration of oxygen in each cylinder was checked with either aThermox probe or a Hersch cell. Percentages of oxygen in the gases used were 100,1-01, 059, 03, 0-1 and 006. In the conditions of our experiments these con-centrations corresponded to oxygen partial pressures of 740, 747, 437, 2'22, 074 and0'44 mmHg. The 'white spot' nitrogen used for anoxic irradiation contained 2 p.p.m.02. Gas was passed through copper tubing to a flowmeter, then through two water-filled Dreschel bottles and finally through the irradiation chamber. The cells wereequilibrated by passing the required gas through the chamber for 5 min at a flow rateof 08 /min. Immediately before irradiation the flow rate was reduced to 0 1 /min toprevent the cells from drying out. When the equilibrating gas contained 0-1 per centoxygen or less the effluent gas from the irradiation chamber was passed through aHersch cell to check that there were no air leaks in the system. Routinemeasurements when nitrogen containing 2 p.p.m. 02 was used gave values of notmore than 20 p.p.m. 02 in the effluent gas after the equilibrium and also after theirradiation period. After each increment of radiation dose three Millipore filters wereremoved from the chamber, which was then resealed and re-equilibrated.Irradiation was continued until the last group of filters had received the requireddose. Unirradiated cells were gassed in the same manner; some for 5 min and othersfor the same length of time as that required to deliver the highest dose of radiation.

2.4. Post-irradiation proceduresEach group of three filters was placed in a universal bottle containing 10 ml of

phosphate buffer which was then shaken vigorously on a Whirlimix for 15 min todislodge the cells. About 90 per cent of the cells could be washed off in this way.

For growth into visible colonies, cells were incubated on pieces of sterilecellophane (50 x 25 mm) lying on the surface of nutrient medium contained in Petridishes. This technique, devised by Alper and Gillies (1958), allows rapid transfer ofcells from one medium to another. To treat cells with penicillin, bacterial suspensionwas placed on pieces of cellophane lying on plates of BAB into which the requiredconcentration of the penicillin in buffer solution had been incorporated before thesterilized medium had set. The bacterial suspension was dispensed on to thecellophane in volumes of 4 /l on each piece by means of an Agla micrometer syringeand was then spread over the surface with an angled glass rod spreader. After pilotexperiments, the concentration of the bacterial suspension could be adjusted so that,whatever treatment was given to the cells, approximately 80 colonies would grow oneach piece of cellophane. A concentration of penicillin of 10 u/ml in nutrient medium(BAB/penicillin) inhibited cell division but caused no killing during a 4 hourincubation at 37°C, whereas after that time significant killing of cells set in abruptly.For that reason the period of incubation with penicillin (always at 10u/ml) waslimited to 4 hours. To terminate exposure of the cells to penicillin, the pieces ofcellophane were transferred at the required times from BAB/penicillin to warmplates of BAB. Unirradiated and irradiated bacteria were treated in the same way, aswere the appropriate control samples which were grown on plates of BAB only.Plates were incubated overnight at 37°C. Surviving fractions were calculated fromthe colony counts.

429

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Fr

eie

Uni

vers

itaet

Ber

lin o

n 11

/13/

14Fo

r pe

rson

al u

se o

nly.

F. . Obioha et a].

2.5. Possibility of oxygen depletion at high dose rateThe PO2 within a cell depends on the supply of oxygen reaching it by diffusion

from its environment. When high dose rates of radiation are used and the ambient pO2is low, significant amounts of oxygen may be used up in radiation chemical reactionsand the radiosensitivity of the cells thereby reduced (Evans 1969). This artefact wasunlikely to affect the results of the experiments reported here, because the bacteriawere exposed in a single layer to the ambient gas. Nevertheless, because a high doserate was needed, it was necessary to check that there was no significant oxygendepletion in the conditions of our experiments. The bacteria were irradiated at doserates of 10, 20, 50 and 100 Gy/min with 01 per cent oxygen in nitrogen as ambientgas. Surviving fractions obtained using these dose rates were not significantlydifferent (figure 1; see §3.1), so a dose rate of 50Gy/min was selected for allsubsequent experiments.

2.6. The determination of o.e.r.The o.e.r. for the killing of the bacteria was calculated from the survival curves

obtained after irradiation of the cells in 100 per cent oxygen and 'white spot' nitrogenunder the conditions of irradiation described above. Results of these irradiations canbe seen in figure 2 (see § 3.2).

2.7. Determination of the K valueAlper and Howard-Flanders (1956) semi-empirically derived their equation

relating partial pressure of oxygen (P) to the extent of radiosensitization caused inbacterial cells; they expressed this in the form

r=(mP+K)/(P+K) (1)

where r is the sensitivity at a particular pressure of oxygen, P, relative to that underanoxia and m is the oxygen enhancement ratio. Sensitivity is defined as the inverse ofthe dose required to give an average of one lethal event per cell, i.e. as Do

1 , Do beingcalculated in the usual way from the slope of the exponential tails of the shoulderedsurvival curves. Howard-Flanders (1958) and Alper (1979) showed that equation (1)could be derived theoretically.

Equation (1) may be rewritten

(r-1)/(m -r)=P/K (2)

Thus if (r- 1)/(m-r) is plotted as a function of P, the graph will be a straight lineprovided the sensitivity of the test system conforms with equation (1). The slope ofthe line gives the value of 1/K (Alper 1976).

An alternative method for calculating K (and also m) makes use of the ratioSmax/S = R for each value of O02 , where Smax is the maximum observable sensitivityat very high pO2 . Then

1/(R- 1) =mP/K(m-1) + 1/(m-1) (3)

(Alper et al. 1967 b). A plot of 1/(R- 1) against P should produce a straight lineenabling values of m and K to be calculated from the intercept and slope respectively.

2.8. Calculation of r and RIn these experiments it was not possible to test the effect of all the gas mixtures on

a particular day so the results were accumulated from 10 separate experiments; all of

430

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Fr

eie

Uni

vers

itaet

Ber

lin o

n 11

/13/

14Fo

r pe

rson

al u

se o

nly.

Alper and Howard-Flanders constants

the experimental points can be seen in figure 1. The survival curve parameters foreach pO2 were calculated from the appropriate pooled results using the Pike-Alpercomputer program (Pike and Alper 1964). The parameters of the individual survivalcurves within each treatment group (i.e. with or without penicillin) fulfilled thestatistical requirements to allow the use of a common extrapolation number, n, so theDo values for each curve were calculated accordingly. These values were used tocalculate r and R.

2.9. Calculation of the effective Do for induction of the penicillin-sensitive lesionWithout penicillin treatment, the bacteria responded to the effects of radiation

plus the enhancing action of oxygen at the relevant PO2. With the treatment, in thepresence of oxygen, there was additional killing due to expression of the penicillin-sensitive lesion. To calculate the sensitivity resulting from expression of the lesion asan isolated effect, we assumed that this was the difference between overall sensitivitywith the penicillin treatment and the sensitivity of the bacteria to radiation when notreatment was given, 'sensitivity' being defined as in § 2.7, i.e.

pD O 1 = oD- 1 _ cD- l (4)

where D0 , pD0 and D0 refer respectively to controls (irradiated but no penicillin),to the induction of the penicillin sensitive lesion alone and to the overall effect.

3. Results

3.1. Survival curvesSurvival curves for controls (i.e. bacteria not incubated in penicillin) are

presented in figures 1 (a) and (b), and the corresponding curves for the overall effect(bacteria incubated in the presence of penicillin) can be seen in figures 1 (c) and (d).The values of cD0 (controls) and o D0 (overall effect) and the common extrapolationnumbers and their 95 per cent confidence intervals are listed in table 1 together withthe derived values of 'D0 ' for the penicillin-sensitive lesion alone, Do .

3.2. Determination of m and KValues of r, (r-1)/(m-r), R and 1/(R-1) are given in table 2. The oxygen

enhancement ratios, measured for control and treated bacteria, were respectivelymC = 329 and mo = 4-82.

In figure 2, values of (r-1 )/(m- r) have been plotted againstpO2 for both controland treated bacteria. Independent values of K and their 95 per cent confidenceintervals, obtained from the slopes of the regression lines were:

K(= KControI) = 459 (4-17-5-03)mmHgKo( =Ktreated) = 440 (3'96493) mmHg

These results were not significantly different and therefore a single regression linewas calculated for all the points, from which

K=444 (419-471)mmHg

Because there was no detectable effect of penicillin on bacteria irradiated in nitrogenit was not possible to obtain a direct value of m for the penicillin-sensitive lesion alone(mp). However, a plot of 1/(R- 1) against P will also give a straight line, according toequation (3) and does not require prior knowledge of the value of m. The intercept onthe axis P=0 gives 1/(m- 1), and the slope gives m/K(m- 1), hence K.

431

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Fr

eie

Uni

vers

itaet

Ber

lin o

n 11

/13/

14Fo

r pe

rson

al u

se o

nly.

432

M

Ct

0n

0

ff

:

(II

F. I. Obioha et al.

00DOSE Gy DOSE Gy

Figure 1. Survival curves of E. coli B/r irradiated in different partial pressures in mmHg ofoxygen, (a) and (b) without penicillin, (c) and (d) with penicillin. (a) and (c): 0, 740;+, 747; *, 222; *, 044; 0, 000. (b) and (d): *, 4-37; x, 074mmHg Broken linesrepresent pO2 740 and zero mmHg transposed from (a) to (b) and from (c) to (d).

IAIAd I

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Fr

eie

Uni

vers

itaet

Ber

lin o

n 11

/13/

14Fo

r pe

rson

al u

se o

nly.

Alper and Howard-Flanders constants

Table 1. D (with 95 per cent confidence intervals) for E. coli B/r incubated with or without penicillin for4 hours after electron irradiation, and the calculated Do for the penicillin-sensitive lesion.

Penicillin-sensitiveControl (no penicillin) Treated (penicillin) lesion alone

pO2 (mmHg) D0 (Gy) 95 per cent CI Do(Gy) 95 per cent CI pD0(Gy)

0 273 259-289 294 259-340044 244 219-277 204 175-245 1244074 198 190-207 159 144-176 807222 159 152-167 130 121-141 713437 133 127-139 104 97-112 4777-47 112 107-117 86 81-93 370

740t 83 80-87 61 56-67 230n = 245 (205-292) n = 137 (090-207)

t 740 mmHg is the PO2 of 100 per cent oxygen saturated with water vapour at 22°C when barometricpressure = 760 mmHg.

Table 2.

Penicillin-sensitiveControl (no penicillin) Treated (penicillin) lesion alone

r-1 1 r-1 1 pO2 (mmHg) r R r R' R

m-r R-I1 m-r R-I R-I

000 1-00 - 329 044 1-00 - 482 0-26 - -044 112 006 294 052 144 013 3.34 043 541 0230'74 1-38 0-20 2-39 0-72 1-85 0-29 2-61 0-62 3-51 0-40222 172 046 1-92 109 226 0-49 213 089 310 048437 2-05 0-85 160 167 2-83 0-92 1-70 1-43 2-07 0937.47 2-44 1-69 1-35 286 342 1-73 1-41 2-44 161 164

740 3-29 - 1-00 - 4-82 - 1-00 - 1'00 -

20

1-5

LE 10

0-5

0-00 2 4 6 8

p 02 mmHgFigure 2. Plot of ) versuspO 2. 0, cells not exposed to penicillin; x, cells exposed to

f(mr penicillin.

433

_ _

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Fr

eie

Uni

vers

itaet

Ber

lin o

n 11

/13/

14Fo

r pe

rson

al u

se o

nly.

F. I. Obioha et al.

R-1R-1

0 2 4 6 8 10p0

2 mmHg

Figure 3. Plot of(R 1 versuspO2 . x, cells not exposed to penicillin; O,cells exposed to

penicillin; 0, values calculated for penicillin-sensitive lesion alone.

Table 3.

m m K K

(Do 100 per cent N2 (eqn. (3)) (eqn. (2)) (eqn. (3))

Do 100 per cent 2

Control 329 348 4-59 442Treated 4-82 432 440 466Penicillin-sensitive

lesion alone - 780 - 5'94

Equations used to derive these values of m and K are shown in parenthesis

In figure 3, values of 1/(R- 1) have been plotted as a function of PO2 for controland penicillin-treated bacteria, together with the values based on calculated values ofPD0 pertaining to the penicillin-sensitive lesion alone. Values of m were obtainedfrom the intercepts, and values of K were calculated from the slopes. All measured orderived values of K and m are given in table 3.

4. DiscussionIn the light of the conclusions of Gillies et al. (1979), Obioha (1981) and Gillies

and Obioha (1982), quoted in the introduction, the value of mp deduced from ourresults may be regarded as the o.e.r. specifically for radiation damage to bacterialmembrane as assessed by a biological test. Although the test of damage to membranewas different from that used by Watkins (1970) or Cramp et al. (1972), our value ofthe o.e.r. agrees well with theirs, and is notably high compared with other tests ofradiation damage to cells.

434

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Fr

eie

Uni

vers

itaet

Ber

lin o

n 11

/13/

14Fo

r pe

rson

al u

se o

nly.

Alper and Howard-Flanders constants

High as it is, the value mp = 8 must be regarded as a lower limit. Its derivationdepends on values of 'PDo' calculated from measured values of cD0 , and oD0 , and istherefore subject to considerably greater error, particularly since the differencesbetween values of CDo 1 and oDo 1 were small at the lower values of PO2 . The mostaccurate determination of 'pD0 ' is 230 Gy, for irradiation with 100 per cent oxygen,and, if mp= 8, 'pD o' under anoxia would be 1840 Gy. This would make Do underanoxia 238 Gy, which is outside the 95 per cent confidence interval for the measuredvalue. We conclude that the various errors involved in the calculation of mp havecombined to give a minimum value. Correspondingly, the derived value, K,= 5'9 mmHg, must be an upper limit.

The results presented diagrammatically in figure 2 provide another and morecogent reason for regarding the true value of Kp as being much nearer to Kc. Figure 2shows (a) that the sensitivities of 'control' and 'treated' bacteria both change withPO 2 in conformity with equation (1), and (b) that the values of Kc and Ko are equal,within experimental error. The response of the penicillin-treated bacteria had twoseparate and resolvable components: that which was identical with the response ofthe control bacteria, and the contribution to overall killing from the attack ofpenicillin on the radiation-induced membrane lesion. When response to radiationhas two components of sensitivity, both of which conform with equation (1) thecombined response will also so conform only if the values of K for the componentsare equal (Alper, in preparation). Similarly, if the combined sensitivity and that ofone of the two components conform with equation (1), and have equal values of K,the second component must itself so conform, and with the same value of K.Determination of the values of Kc and K, by the use of equation (2), was subject tomuch less error than was the determination of Kp by use of equation (3). The valuesof K. and Ko might have been slightly different, as the confidence limits show, andthis would permit Kp also to be slightly greater than Kc; but it could not have beenvery different.

It is plausible that the parameters Kp and mp should apply also to other radiation-induced membrane lesions in E. coli B/r, since the constants of equation (1) aredetermined by the chemical nature of the structure within which energy is deposited,together with the concentration of the reactive species in their immediate vicinityand their reaction rates (Howard-Flanders 1958, Alper 1979). Howard-Flandersproposed that free-SH groups compete with oxygen in the metionic reaction. Thishas been well supported by results of Michael and Harrop (1979) with bacteria andCullen et al. (1980) with mammalian cells. Values of K for cell killing of cells of bothclasses were correlated with the cellular concentration of non-protein-SH.

Values of m and K are interdependent, when specific sites of energy depositionare under consideration (Alper 1979). It is therefore improbable that, for two suchdifferent cellular components as membranes and DNA, values of K should be nearlythe same, but values of m very different. Yet that would follow if, as is commonlysupposed, lethal effects of radiation were due only to energy deposition events in thebacterial DNA, i.e. if the parameters K, and m~ were associated entirely with thoseevents.

The techniques used in our experiments made it possible to determine values ofm and K applicable to a test of biological damage specifically to bacterial membrane;but no comparable technique exists by which those parameters, or even m alone, canbe unequivocally determined for biological damage consequent on energy depositionspecifically in the bacterial DNA. Indeed, the weight of the evidence is against an

435

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Fr

eie

Uni

vers

itaet

Ber

lin o

n 11

/13/

14Fo

r pe

rson

al u

se o

nly.

F. I. Obioha et al.

o.e.r. of much more than one for such events. Values of o.e.r. substantially greaterthan one for some tests of biological damage to DNA have indeed been observed(e.g. Hutchinson and Arena 1960), but this has always involved irradiation of wholecells, so that DNA was attached to membrane during irradiation. Several results ofthe irradiation of DNA extracellularly have yielded values of o.e.r. of one or less (e.g.Ephrussi-Taylor and Latarjet 1955, van der Schans 1973), and this has beenexplained as attributable to the want of sufficient SH-containing compounds in thetest tube experiments (e.g. Howard-Flanders 1960, Hutchinson 1961). But thereare various observations that cannot be reconciled with the assumption thatbiological damage to intracellular DNA is enhanced to any significant extent by thepresence of oxygen. If it were, values of o.e.r. should be high for the killing of cells inwhich lesions in DNA are particularly effective by virtue of deficiency in capacity torepair DNA; but the reverse is true. It has been shown that values of o.e.r. for thekilling of bacteria are inversely correlated with the radiosensitivity of geneticallyrelated strains (Alper 1968). Values of o.e.r. have been found to be one for very U.V.-sensitive mutants, like UVSI of Chlamydomonas reinhardii (Davies 1967) or E. coliBs 12 (Forage and Alper 1970). Oxygen has been found to be without effect on theextent of radiation-induced damage to two different functions of yeast DNA (Kieferet al. 1980, Weber and Kiefer 1983).

It would follow that values of K observed for the killing of cells are, in fact,determined by the values of K proper to lethal lesions in membrane.

On the basis of that hypothesis, it is possible to calculate the relativecontributions to the killing of E. coli B/r from primary energy deposition inmembrane and DNA. Let the overall sensitivity of the bacteria to killing underanoxia be Sc, made up of SD (DNA) and SM (membrane). Let the o.e.r. observed forcell killing be m (= 3'3, from our results) and let the o.e.r. for membrane damage bemM (=mp=8, from our results).

Sc = SD + SM

For irradiation in oxygen,

meSh= lSD+8 SMmS) = 1SD + 8 M

3 3(SD+Sm)=SD+8SM

i.e. SD/SM'2.Thus for irradiation under anoxia, about two-thirds of the lethal events come

from primary energy deposition in DNA; but with full oxygenation, the correspond-ing contribution is reduced to about one-fifth.

AcknowledgmentsIt is a pleasure to thank Dr W. A. Cramp and Dr N. T. S. Evans for their advice.

One of us (F.I.O) is grateful to the Federal Scholarship Board of the NigerianGovernment for financial support.

ReferencesALPER, T., 1968, Nature, Lond., 217,862; 1979, Cellular radiobiology (Cambridge: Cambridge

University Press).ALPER, T., and GILLIES, N. E., 1958, J. gen. Microbiol., 18, 461.ALPER, T., and HOWARD-FLANDERS, P., 1956, Nature, Lond., 178, 978.ALPER, T., MOORE, J. L., and BEWLEY, D. K., 1967a, Radiat. Res., 32, 277.

436

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Fr

eie

Uni

vers

itaet

Ber

lin o

n 11

/13/

14Fo

r pe

rson

al u

se o

nly.

Alper and Howard-Flanders constants

ALPER, T., MOORE, J. L., and SMITH, P., 1967b, Radiat. Res., 32, 780.BLUMBERG, P. M., and STROMINGER, J. L., 1974, Bacteriol. Rev., 38, 291.BRAUN, V., and SIEGLIN, U., 1970, Eur. J. Biochem., 13, 336.COSTERTON, J. W., INGRAM, J. M., and CHENG, K.-J., 1974, Bacteriol. Rev., 38, 87.CRAMP, W. A., WATKINS, D. K., and COLLINS, J., 1972, Int. J. Radiat. Biol., 22, 379.CULLEN, B. M., MICHALOWSKI, A., WALKER, H. C., and REVESZ, L., 1980, Int. J. Radiat. Biol.,

38, 525.DAVIES, D. R., 1967, Nature, Lond., 215, 829.EPHRUSSI-TAYLOR, H., and LATARJET, R., 1955, Biochim. biophys. Acta, 16, 183.EVANS, N. T. S., 1969, Radiat. Effects, 1, 19.FORAGE, A., and ALPER, T., 1970, Int. J. Radiat. Biol., 17, 527.GILLIES, N. E., and OBIOHA, F. I., 1982, Int. J. Radiat. Oncol. Biol. Phys., 8, 453.GILLIES, N. E., OBIOHA, F. I., and RATNAJOTHI, N. H., 1979, Int. J. Radiat. Biol., 36, 587.HOWARD-FLANDERS, P., 1958, Adv. Biol. Med. Phys., 6, 553; 1960, Nature, Lond., 186, 485.HUTCHINSON, F., 1961, Radiat. Res., 14, 721.HUTCHINSON, F., and ARENA, J., 1960, Radiat. Res., 13, 137.KIEFER, J., GOCKE, E., and KOESTER, K., 1980, Int. J. Radiat. Biol., 38, 267.MICHAEL, B. D., and HARROP, H. A., 1979, Radiation sensitizers, edited by L. W. Brady

(New York: Masson Publishing U.S.A. Inc.), p. 14.OBIOHA, F. I., 1981, PhD Thesis, University of London.PIKE, M., and ALPER, T., 1964, Br. J. Radiol., 37, 458.VAN DER SCHANS, G. P., 1973, cited by Blok and Loman, Curr. Topics Radiat. Res., 9, 165.WATKINS, D. K., 1970, Adv. Biol. Med. Phys., 13, 289.WEBER, K. J., and KIEFER, J., 1983, Proc. 7th Int. Congress Rad. Res. B3-32 (Amsterdam:

Martinus Nijhofl).

437

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Fr

eie

Uni

vers

itaet

Ber

lin o

n 11

/13/

14Fo

r pe

rson

al u

se o

nly.