Proc. Nat!. Acad. Sci. USAVol. 85, pp. 2553-2557, April 1988Biochemistry

Structure of the octopine synthase upstream activator sequence(Agrobacterium tumefaciens/transcriptional activator)

ScoTT M. LEISNER AND STANTON B. GELVIN*Department of Biological Sciences, Purdue University, West Lafayette, IN 47907

Communicated by H. Edwin Umbarger, December 31, 1987 (received for review November 9, 1987)

ABSTRACT We have identified a transcriptional activat-ing element within the 5' flanking sequence of the Agrobacte-rium tumefaciens octopine synthase (ocs) gene that is necessaryfor ocs expression in transformed tobacco calli. This element islocated between 333 and 116 base pairs upstream from thetranscription initiation site and functions independent of ori-entation when placed upstream of the ocs gene. It does notfunction in either orientation when placed downstream of thegene, nor can it activate its promoter when separated by adistance of 608 base pairs. Deletion analysis indicates thatsequences essential for activator function are localized between222 and 177 base pairs upstream of the transcription initiationsite. Another region, located between 333 and 249 base pairsupstream of the transcription initiation site, does not as amonomer activate the ocs promoter, but it can as a dimer.

Transcriptional control regions of eukaryotic genes consistof three major classes of sequences: (i) proximal controlsignals, (ii) distal control signals, and (iii) activating se-quences (1-3). Proximal control signals, such as the"TATA" box, are located -30 base pairs (bp) upstream ofthe transcription initiation site. Distal control signals, suchas the "CAATT" box, are located 50-90 bp upstream of thetranscription start. Activating sequences, such as transcrip-tional enhancers and the upstream activating sequences ofyeast (4, 5), are usually located >100 bp upstream of thetranscription start site. These sequences function in bothorientations upstream of homologous and heterologous pro-moters to enhance transcription. Many activators can alsoconfer tissue specificity or inducibility on promoters (1-3).Although many activator sequences have the ability tofunction over long distances, only enhancers stimulate tran-scription when placed downstream of genes.

Transcriptional activators are composed of short con-served sequence motifs that serve as transcription factorbinding sites (1-3, 6). Two common motifs are the simianvirus 40 (SV40) enhancer core GTGGAWW (where W iseither A or T) and alternating stretches of purines andpyrimidines (with the potential to form Z-DNA).Workers in our laboratory have been interested in the

transcriptional regulation of genes transferred from Agro-bacterium tumefaciens into plant cells, where they areexpressed (7). One of these is the octopine synthase (ocs)gene that is constitutively expressed in a wide variety ofplant tissues (8). This gene has a TATA box situated 32 bpupstream of the transcription initiation site (Fig. 1) (9, 10). Inaddition, sequences located between 170 and 294 bp up-stream from the transcription start are also required foroctopine synthase activity in tobacco crown gall tumors (11).This region contains two sequence motifs that are homolo-gous to the SV40 enhancer (6), including an SV40 enhancercore sequence (GTGGAAA) and a 10-bp stretch of alternat-ing purines and pyrimidines capable of forming Z-DNA (Fig.

1). The presence of these motifs in an upstream regionrequired for transcription of the ocs gene suggested to us thatthe nucleotides between - 170 and -294 may contain atranscriptional enhancer-like element. In this report, weshow that the ocs element activates its own promoter in amanner similar to other activator sequences. We also showthat the ocs activator has a complex structure.

MATERIALS AND METHODSMaterials. Restriction endonucleases were purchased

from Bethesda Research Laboratories and used according tothe manufacturer's specifications. DNA polymerase I Kle-now fragment and T4 DNA ligase were obtained fromPharmacia and used according to the manufacturer's speci-fications. [a-32P]dCTP was purchased from Amersham.DNA fragments were labeled with an Amersham nick-translation kit. Reagents for the octopine synthase assay andantibiotics were purchased from Sigma. Reagents for deter-mining protein concentrations were purchased from Bio-Radand used according to the manufacturer's specifications.

Strains and Culture Conditions. Escherichia coli strainswere grown in LB medium (13). Agrobacterium strainLBA4404 (14) was grown in AB minimal medium (15).Antibiotic concentrations, when used, were for E. coli:kanamycin, 50 pug/ml; ampicillin, 100 ,ug/ml; for A. tumefa-ciens: kanamycin, 100 tug/ml; rifampicin, 10 ,ug/ml. Theseexperiments were conducted under P1 containment condi-tions as specified by the National Institutes of HealthRecombinant DNA Guidelines.

Construction of Plasmids to Test the Effects of Position andOrientation on the ocs Gene Activator. The ocs structuralgene is located within a 2.57-kbp BamHI/Sma I subfragment[base pairs 13774-11207 (10)] of BamHI fragment 17 [basepairs 13774-9062 (10)] from the Ti plasmid pTiA6. This2.57-kbp fragment was cloned into the BamHI and Sma Isites of the binary plasmid pBinl9 (16) to yield the plasmidpOCSA2 (Fig. 2B). This plasmid contains 116 bp upstreamfrom the transcription start, the entire ocs structural gene,and -1200 bp downstream from the polyadenylylation site.The activator was isolated as a 217-bp Acc I/BamHI

subfragment [base pairs 13991-13774 (10)] of BamHI frag-ment 2 (10). This fragment contains the sequence from - 333to -116 bp relative to the ocs mRNA initiation site (9). TheAcc I site was converted to a BamHI site by linker attach-ment (13) and the fragment was cloned, in both orientations,into the unique BamHI site upstream of the ocs gene inpOCSA2. Alternatively, the ends were filled in with Klenowfragment and the resulting fragment was cloned into theunique Sma I site downstream of the ocs gene (Figs. 2 and3). Cloning the activator sequence upstream of the ocs genein the correct orientation (pEN1) reconstitutes the ocs 5'flanking sequence to - 333 (Fig. 3). pENRl contains the

Abbreviations: SV40, simian virus 40; CAT, chloramphenicol ace-tyltransferase.*To whom reprint requests should be addressed.

2553

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 7,

202

0

2554 Biochemistry: Leisner and Gelvin

A

-427 -402 - 264 -253-294

CORE

-183 -174 +1 +27

170 -32 -25 lZZ}F-II~IATCK---ACCAAA ATGZ- DNA TATA BOX

B- 333ACAGGCCAAA TTCGCTCTTA GCCGTACAAT ATTACTCACC GGTGCGATGC CCCCCATCGT

AGGTGAAGGT GGAAATTAAT GATCCATCTT GAGACCACAG GCCCACAACA GCTACCAGTTcore

pal indromeTCCTCAAGGG TCCACCAAAA ACGTAAGCGC TTACGTACAT GGTCGATAAG AAAAGGCAAT

Z - DNA

TTGTAGATGT TAACATCCAA CGTCGCTTTC AGGGATCCTT TTTACCGACA ACTCATCCAC

ATTGATGGTA GGCAGAAAGT TAAAGCATTA TCGCAAGTCA ATACTTGCCC ATTCATTGAT

CTATTTAAAG GTGTGGCCTC AAGGATAATC GCCAAACCAT TATATTTGCA ATCTACCAA

+1ATG+27

FIG. 1. (A) Structure of the 5' flanking sequence of the octopine synthase gene (9, 10). Positions of various elements flanking the ocsstructural gene are shown. El, Right T-DNA border. The open box between - 294 and - 170 contains a region of the ocs 5' flanking sequenceessential for expression in tobacco tumors (11). Shaded areas within this region show the location of the SV40 enhancer core sequence and thenucleotides capable of forming Z-DNA. The locations of the TATA box, the transcription initiation site, and the translation initiation site arealso shown. (B) Nucleotide sequence of the 5' flanking region of the ocs gene. The sequence starts at position - 333 and ends at the translationinitiation site (ATG). The transcription initiation site (+ 1) is marked by an arrow. The SV40 enhancer core and the putative Z-DNA sequencesare underlined singly. The 16-bp palindrome described by Ellis et al. (12) is overlined.

activator in the inverted orientation at the 5' end, pEN2contains the activator in the native orientation but down-stream of the ocs gene, pENR2 contains the activator in theinverted orientation downstream of the ocs gene, andpENR22 contains a tandem duplication of the activator inthe inverted orientation downstream of the ocs gene (Fig. 3).

RB

RB

FIG. 2. The intermediate plasmids pOCSA1 (A) and pOCSA2(B). These plasmids are derived from the binary plasmid pBinl9(16). The circles are a schematic map of pBinl9. The kanamycin-resistance gene controlled by the nopaline synthase promoter isdesignated Kmr; LB, left T-DNA border; RB, right T-DNA border.Box represents the ocs structural gene. The TATA element for theocs gene is represented as a solid box. B, BamHI; D, Dra I; H,HindIII; N, Nco 1; S, Sma I; Ss, Sst I. (A) Intermediate plasmidpOCSA1. This plasmid was constructed by cloning the 1.5-kbp DraI fragment into the Sma I site of pBinl9. (B) The intermediateplasmid pOCSA2. This plasmid was constructed by cloning the2.57-kbp BamHI/Sma I fragment into the BamHI and Sma I sites ofpBinl9.

The 2.57-kbp BamHI/Sma I fragment harboring the ocsgene was cleaved in the TATA box of the ocs promoter and160 bp downstream from the polyadenylylation site by Dra I(Fig. 2). This 1.5-kbp Dra I fragment [base pairs 13685-12221(10)] was cloned into the unique Sma I site of pBinl9 (16) toyield the plasmid pOCSA1 (Fig. 2A). HindIl linkers wereattached to the 217-bp activator and the resulting fragmentwas cloned into the HindIII site of pOCSA1, in the correctorientation, to yield the plasmid pENA1 (Fig. 3). pENA1contains the entire activator and coding sequences of the ocsgene but has a truncated promoter (Fig. 3).A 608-bp HindIII/Bgl II fragment [base pairs 6631-6023

(10)] from EcoRI fragment 7 [EcoRI fragment B; base pairs5545-12823 (10)] was cloned into the unique HindIlI andBamHI sites of pOCSA2 to give the plasmid pSPOC4. Thisfragment is derived from the protein-coding region of thetmsl gene (10, 17). HindIII linkers were attached to the217-bp activator fragment, which was subsequently clonedinto the HindIII site of pSPOC4 in the correct orientation toyield the plasmid pSPEN0.6 (Fig. 3). In this plasmid, 608 bpseparate the activator from the ocs promoter. The activatorwith HindIII ends also was cloned into the HindIII site ofpOCSA2 giving rise to plasmids in which the activator is inthe correct (pEN4) or inverted (pENR4) orientation. In theseconstructions, the activator and the promoter are separatedby 30 bp.

Construction of Plasmids Harboring Deletions of the Octo-pine Synthase Activator. The 217-bp activator fragment wascleaved with various restriction endonucleases, BamHI link-ers were added, and the resulting fragments were clonedindividually into the BamHI site of pOCSA2, yielding thedeletions shown in Fig. 4. The two Sau3A fragments (Saul33and Sau84) were cloned without linker attachment since theyhad BamHI-compatible cohesive ends.

Conjugation of the Constructs into A. tumefaciens. Plasmidswere conjugated into A. tumefaciens strain LBA4404 by a

Proc. Nati. Acad. Sci. USA 85 (1988)

Dow

nloa

ded

by g

uest

on

Dec

embe

r 7,

202

0

Proc. Natl. Acad. Sci. USA 85 (1988) 2555

N< - NUn - N EtU Z Z Z Z

INC. ° ui L ui LTIME rF r7 n7(HRS) o 2 0 2 2 2 0 2 0 2

__--qmmw-mj~~ARG --

(0N dN - 't z

z z Z Z a.aL a L XL CL

n ra n n0 2 0 2 0 2 0 2 0 2TTTT

j Ro I_

OCT A

pOCSA2

pEN

pENR

pEN2

pENR2

pENR22

OCS Activity

Ip ocs- -

c--- +

P°s -1 +

pi ocs- +t+-p ocs-a.I -

pENA. I ocs;pEN4 p

30" bp

pENR 4 PI ocs-30 bp

pSPENO0.6 bp

608 bp

IOcs- |. ocs gene;[bEocs promoter; E1I ocs activator

o 0 0 aN q N 0UCU

0 CDj OD No-~ OD r()

FIG. 3. The ocs activator sequence functions in both orientations at the 5' end of the ocs gene. Electrophoretogram of octopine synthaseassays performed on calli incited with the constructions noted above. Positions of octopine (OCT) and arginine (ARG) are marked. Numbersabove indicate the time of incubation in hours. Origin (+) is at the bottom of the photo. A schematic diagram is shown of the plasmids pOCSA2,pEN1, pENR1, pEN2, pENR2, pENR22, pENAl, pEN4, pENR4, and pSPENO.6 described in the text. El, ocs promoter (-116 to -1);[oca 1, ocs gene (+ 1 to + 1273) that includes the 5' untranslated sequence, the ocs structural gene, and the 3' untranslated sequence. Blankspace after arrow represents the 1153 bp downstream of the polyadenylylation site; i<, ocs activator (-333 to -116). Thirty base pairsseparates the ocs activator from the ocs promoter in the constructs pEN4 and pENR4, and 608 bp separates the activator and promoter inpSPENO.6. Figure is not drawn to scale. Twenty calli were analyzed for ocs activity for each construction. The octopine in the spots shownis approximately 6000 pmol for pEN1, 6200 pmol for pENR1, 6500 pmol for pEN4, 6000 pmol for pENR4, and 0 pmol for all of the remainingconstructions. Octopine standards used for quantitating the levels of octopine synthesized by extracts from calli incited by variousconstructions are shown below. The amount (pmol) of octopine present within each spot is indicated.

triparental mating (16) involving the E. coli strain harboringthe recombinant plasmid, E. coli strain MM294/pRK2013(18), and A. tumefaciens strain LBA4404. The Agrobacte-rium transconjugants were selected on AB minimal platescontaining 0.5% glucose, rifampicin (10 iug/ml), and kana-mycin (100,ug/ml).Transformation of Plants and Octopine Assays. Tobacco

leaf disks were infected with the Agrobacterium strainsharboring recombinant plasmids by the method of Horsch etal. (19). Kanamycin-resistant calli were assayed for octopinesynthase activity by the method of Otten and Schilperoort(20). Twenty different calli were assayed for each construc-tion. Extracts contained 20-30 pAg of protein as determinedby the method of Bradford (21). Constructs were denoted asoctopine synthase-positive when >40%o of the calli analyzedexhibited detectable octopine synthase activity. Construc-tions denoted as octopine synthase-negative never inducedcalli that exhibited detectable octopine synthase activity. S1nuclease analysis of pEN1, pENR1, and Sau84 dimer octo-pine synthase mRNA initiation sites was performed asdescribed by Gelvin et al. (22) with strand-specific probesmade from an M13 template (23).

RESULTS

The Upstream Region of the ocs Gene Contains a Transcrip-tional Activator Sequence. To test whether the sequence

between - 294 and - 170 functions as an enhancer of the ocs

promoter, the constructions shown in Fig. 3 were madeusing the binary plasmid pOCSA2 (Fig. 2). After mobiliza-tion of these plasmids into A. tumefaciens strain LBA4404,tobacco leaf disks were infected and kanamycin-resistantcalli were selected (19). These calli were assayed for octo-pine synthase activity (20).

Calli containing the construct pOCSA2, that lacks the ocsupstream element, had no detectable octopine synthaseactivity (Fig. 3). When the upstream element was clonedupstream of the ocs gene in the correct orientation (pEN1),reconstructing the octopine synthase upstream region, octo-pine synthase activity was restored. Cloning the sequenceupstream of the ocs gene in the opposite orientation(pENR1) also resulted in octopine synthase activity. S1nuclease protection experiments indicated that the tran-scripts from pEN1 and pENR1 start from the correct tran-scription initiation site (data not shown). Together theseresults show that the ocs upstream sequences activate theocs promoter independent of orientation.When the activator sequence was cloned downstream of

the ocs gene in either the correct (pEN2) or reverse (pENR2)orientation and introduced into tobacco calli, octopine syn-thase activity was not detected (Fig. 3). Similarly, when twocopies of this element were cloned downstream of the ocsgene (pENR22), or when the activator was moved 1 kbpcloser to the 3' end of the ocs gene (data not shown),

+

aN(D

Biochemistry: Leisner and Gelvin

Dow

nloa

ded

by g

uest

on

Dec

embe

r 7,

202

0

2556 Biochemistry: Leisner and Gelvin

INC.TIME(HRS)

AR G-

r(r) (D0Or0

020(n <x0 2 0 2

I

(D

0U)

0 2 C

OCT-

BamHI Sau3a A uIpENI I I

-333 -249 -222C

Core

RsoI

-177

Z - DNA

Sou 133Al u 106 1 -

Rsa6I -Sau84 - ( copy )Sau84 - ( 2 copies )

FIG. 4. Deletion analysis of the ocs acgram of octopine synthase assays performedconstructs shown. The positions of octopi(ARG) are marked. Numbers above indicatin hours. Origin (+) is at the bottom of tendonuclease map of the ocs activator is siof the activator cloned into the BamHI site

is, ~ normal position to a position 30 bp further upstream of theocs promoter (pEN4 and pENR4), it still functioned in both

co CD orientations. However, when a 608-bp fragment was insertedc, = between the activator and the promoter of the ocs gene,

enCo octopine synthase activity was abolished. These resultsm m agree with the findings of De Greve et al. (9), who showed2 0 2 that an insertion of a 2500-bp fragment into the BamHI site's (which lies between the ocs activator and promoter) resulted

in the loss of octopine synthase activity. These data indicatethat the ocs activator is sensitive to its distance from the ocspromoter, although it is possible that the spacer containssequences that inhibit transcription.

Structure of the Octopine Synthase Upstream ActivatorSequence. To determine which sequences are necessary foractivator function, various restriction endonuclease frag-ments of the ocs activator were cloned into pOCSA2 (Fig. 4).When the Saul33 fragment (-249 to - 116 from the ocstranscription initiation site) was cloned upstream of the ocsgene, octopine synthase activity was observed. This frag-ment does not contain the SV40 enhancer core sequence.Octopine synthase activity was also observed when the

BamHI OCS Activity AIulO6 fragment (- 222 to - 116) was cloned upstream of theocs- + ocs gene. No activity was observed, however, when the-116ocgeewaRsa61 fragment (- 177 to -116) was cloned onto the ocs

--+ocs- + gene. Thus, a 45-bp fragment (-222 to -177 from the-+OCS- + transcription initiation site) contains an element necessaryocs- + for the expression of octopine synthase activity in tobaccoMocs- - calli. Recent experiments indicate that this 45-bp fragment,

Focs- + when cloned into the BamHI site of pOCSA2, can activatethe ocs promoter (data not shown). When the Sau84 frag-

tivator. Electrophoreto- ment (- 333 to - 249) was cloned onto the ocs gene as aI on calli incited with the tandem dimer, octopine synthase activity was observed. S1ine (OCT) and arginine nuclease protection experiments indicated that the tran-.e the time of incubation scripts from this construct start from the correct initiation

hown below. Fragments site (data not shown). However, a monomer of the Sau84of pOCSw.2 are shown. fragment failed to stimulate octopine synthase activity.

Numbers below the restriction sites indicate their position relativeto the transcription initiation site. The Acc I site was converted to aBamHI site as described. The location and orientation of the ocsgene are indicated. The locations of the SV40 enhancer core and theputative Z-DNA sequences are indicated. Twenty calli were ana-lyzed for each construction. The octopine present within the spotsshown is approximately 6500 and 8000 pmol for Saul33 incubatedfor 0 and 2 hr, respectively, 3700 and 6800 pmol for Alu106incubated for 0 and 2 hr, respectively, 6400 and 7900 pmol for theSau84 dimer incubated for 0 and 2 hr, respectively, and 0 for Rsa6land the Sau84 monomer.

octopine synthase activity was not detected. Therefore, theocs upstream element does not greatly stimulate octopinesynthase activity downstream of the ocs gene. Alternatively,the lack of stimulation of ocs gene activity could also beexplained by a distance effect (see below). In our assaysystem, the limit of detection of octopine is -30 pmol(however, it is difficult to photograph spots of <300 pmol).By comparing the spot intensity from calli harboring con-structions designated as octopine synthase positive withknown standards, 2500-8100 pmol of octopine was synthe-sized during a 2-hr incubation.The ocs activator may function in both orientations up-

stream of the ocs gene because it is a bidirectional promoter.To test this hypothesis, a plasmid was constructed (pENAl)containing the activator but lacking the ocs promoter (thosesequences between - 116 and -28, including the TATAelement). Octopine synthase activity was not detected incalli harboring this construct (Fig. 3). Activator function istherefore dependent on promoter-proximal functions.To test whether the ocs activator can function over a long

distance from the ocs promoter, the plasmids shown in Fig.3 were constructed. When the activator was moved from its

DISCUSSIONThe upstream region of the octopine synthase gene containsa transcriptional activator element that activates the ocspromoter independent of orientation. However, the ocselement does not stimulate the ocs promoter when placeddownstream of the ocs gene. If this is not simply due to adistance effect, then the ocs activator is unlike classicalanimal viral and cellular enhancers and functions more likethe upstream activating sequences of yeast (4, 5).

Ellis et al. (12, 24) showed that sequences upstream of theocs gene can activate a heterologous promoter. The chlor-amphenicol acetyltransferase (CAT) gene under control ofthe maize alcohol dehydrogenase (ADHI) promoter is poorlyexpressed in tobacco. When the ocs activator was cloned, ineither orientation or at either end of the CAT gene, CATactivity was expressed at high levels in tobacco. The ocsactivator therefore could stimulate the ADHJ promoter. Thediscrepancy between the data of Ellis et al. (12) and our dataconcerning the effectiveness of the ocs activator down-stream of a gene may be explained in at least three ways.First, our octopine synthase assay is less sensitive than theCAT assay used by Ellis et al. (12, 24). Second, it is possiblethat different factors affect transcription in the transientexpression system used by Ellis et al. (12) and our stableexpression system. Finally, the ocs upstream element mayactivate the ocs and ADHI promoters in different ways. It isknown that the SV40 enhancer can activate various promot-ers to different extents (25).Only a few transcriptional activator sequences have been

identified in genes expressed in plants, including the ribu-lose-1,5-bisphosphate carboxylase small subunit (rbcS)genes (26, 27), the chlorophyll a/b binding protein (cab)

i.fts;

Proc. Natl. Acad. Sci. USA 85 (1988)

::

Dow

nloa

ded

by g

uest

on

Dec

embe

r 7,

202

0

PULMONARY FUNCTION AND CHEST IRRADIATION

in any animal. These findings indicate that theincreased rigidity of the lungs-chest resultedfrom changes within the lung. Three of theseanimals were restudied 172, 228 and 273 days,respectively, following onset of irradiation. Totaland lung compliance was further decreased in allthree. Thoracic cage compliance was below thepre-irradiation value in only one of the three(Figure 2).One nonirradiated and two irradiated dogs were

studied immediately before and after opening thechest. The values obtained with the chest opencompared favorably with those with the chestclosed (Figure 2).

Diffusing capacityDiffusing capacity did not change consistently

in three animals studied two weeks and four

TABLE I

Values for several measures of pulmonary function beforeand at various time intervals after the onset of

fractional irradiation to the chest

Weight TimeDog and of Min.Do. sex study FRC vent. pCO2 p02 Dco

mi. COL./min. min./ /

days ml. STPD mm. Hg mm. Hg mm. Hg1 38.5 lbs. C* 650 2.5 45.6 74.5 5.0

Male 80 609 4.2 51.9 71.1 2.3126 490 1.8 20.5 65.2 1.5

2 49 lbs. C 678 4.7 48.7 82.9 3.7Male 104 615 1.9 54.6 68.5 3.4

161 492 2.9 48.1 47.1 1.57 45 lbs. C 754 2.4 45.6 90.0 3.7

Male 97 269 2.0 38.3 84.2 3.7147 332 3.8 28.9 69.9 2.0

8 32 lbs. C 3.4 31.6 101.1 3.5Male 74 3.4 35.9 91.9 3.6

141 465 2.2 44.4 95.0 2.812 48 lbs. C 2.1 43.5 77.7 4.4

Male 87 566 3.9 30.3 86.2 3.1138 546 2.8 33.1 99.5 2.4273 344 3.9 33.5 81.7 3.6

13 46.5 lbs. C 7.2 38.3 84.4 5.7Male 64 545 2.3 41.6 77.7 3.1

104 509 1.9 47.9 80.7 2.0228 388 9.7 50.1 33.9 1.0

14 41 lbs. C 519 4.3 33.0 81.1 3.9Female 70 2.3 51.3 92.4 3.0

140 513 2.9 34.6 89.6 2.0172 418 3.1 53.0 80.6 4.1

Average C 3.8 40.9 84.5 4.382 2.9 43.4 78.9 3.2137 2.6 36.8 78.1 2.0

* C, control.

: 1'1600. 1

:461200 :e. 11000

600 ..'

400 1'

200. lxCage

0 10 1S

AP Cm.HsO Lung10 10 50 40 50550 0to 20 30 40

FIG. 2. RELATION BETWEEN PRESSURE AND VOLUMEIN CAGE, LUNGS AND TOTAL THORAX OF DOG No. 13 BE-FORE AND AT THREE PERIODS AFTER FRACTIONAL IRRADI-ATION TO THE CHESTOn the two hundred twenty-eighth day, values for the

lung were also obtained immediately after opening thechest. Each point represents the average of three meas-urements.

weeks following irradiation with a single dose.It decreased progressively in those animals stud-ied more than two months following onset offractional irradiation (Table I, Figure 3). Controlvalues from seven animals averaged 4.3 ml. perminute per mm. Hg. By 82 days following theonset of irradiation, diffusing capacity was be-low the control value in five, above it in one andthe same in one. The average value was 3.2 ml.

_6.0

2.0-

40

103.

1.0

20 40 60 80 100 120 140 160Days

FIG. 3. CHANGE OF PULMONARY DIFFUSING CAPACrrYIN SEVEN DOGS FOLLOWING FRACTIONAL IRRADIATION TOTHE CHESTEach point represents one determination.

589

Dow

nloa

ded

by g

uest

on

Dec

embe

r 7,

202

0