Proc. Nati Acad. Sci. USAVol. 78, No. 7, pp. 4151-4155, July 1981Biochemistry

Revertant seedlings from crown gall tumors retain a portion of thebacterial Ti plasmid DNA sequences*

(tumor reversal/Agrobacterium tunmfaciens/genetic engineering of plants)

FUNMEI YANGt AND ROBERT B. SIMPSONDepartment of Microbiology and Immunology, University of Washington, Seattle, Washington 98195

Communicated by Armin C. Braun, March 30, 1981

ABSTRACT BT37 is a crown gall teratoma incited on tobaccoby Agrobacterium tumefaciens containing pTi-T37, a nopaline-type Ti plasmid. Treatment of this cloned tumor tissue with kinetinat 1 mg/liter results in the formation of relatively normal-ap-pearing shoots. These shoots can be induced to root and set viableseed. In contrast to BT37 tissue, the derived tissues are not phy-tohormone independent and do not produce nopaline. The re-verted plants, like normal tobacco plants, are susceptible to in-fection by A. tumefacien. This loss oftumorous traits is accompaniedby the loss of most of the Ti plasmid sequences (T-DNA) found inBT37 DNA. Southern blot analysis indicates that the revertanttissues have lost the central portion of the T-DNA, which containsthe "common DNA" sequences, a highly conserved region of theTi plasmid that has been found to be incorporated into all tumorsstudied. Thus, these sequences appear necessary for oncogenicityand tumor maintenance and their loss is probably directly relatedto tumor reversal. The reverted plants as well as the plants ob-tained from seed, however, do retain sequences homologous to theends of the T-DNA present in the parental teratoma. The per-sistence of foreign DNA sequences during the process of meiosisand seed formation has important implications for the possibilityof the genetic engineering of plants.

The introduction of foreign DNA into normal cells can trans-form them into tumor cells. These cells can partially or com-pletely recover from their tumorous state by a process knownas tumor reversal or reversion. In this communication we de-scribe the phenotypic and molecular events accompanying thereversal of a plant tumor, crown gall.

The inciting organism ofcrown gall tumors is a soil bacterium,Agrobactertium tumefaciens, which contains a large tumor-in-ducing (Ti) plasmid (1-4). Properties of these tumors are: (i) incontrast to normal cells, tumor cells can proliferate autorto-mously in axenic culture without an exogenous supply of plnthormones, auxin and cytokinin (5); (ii) crown gall tumors, butnot normal cells, generally synthesize unusuaLamino acid de-rivatives called opines (6-8); and (iii) a portion of the plasmidDNA from the inciting bacteria (T-DNA) is incorporated stablyin the nucleus oftumor cells attached to plant DNA (9-18). Thefunction of this T-DNA in tumorigenesis is not yet clear, butit is transcribed (16, 19-22) and probably translated (23) in tu-mor cells. Although the amount ofT-DNA found in various tu-mor lines is different, all include a "core" sequence (12, 13)which is probably responsible for oncogenicity and maintenanceof the transformed state. A portion of this core T-DNA is ho-mologous to the "common DNA" sequences, which are highlyconserved among diverse types of Ti plasmids (24, 25). In ad-dition, deletions and insertions within the T-DNA can affect

virulence (26-30), supporting the hypothesis that the T-DNAencodes the "oncogenes."

Normal-appearing plants have been regenerated from bothunorganized (31-34) and teratoma-type (35, 36) crown gall tu-mors. Some tissues from shoots regenerated from a teratomastill exhibited tumorous traits (35, 36) and retained T-DNA (11,13, 37). The parts of this shoot that had gone through meiosisno longer exhibited tumorous traits (36) and lacked detectableT-DNA (11, 13, 37). Also, spontaneous revertants from the sametumor had normal morphology and lacked T-DNA (36, 37).

In this study we examined the T-DNA content of revertantplants obtained from a cloned teratoma tumor by hormonetreatment. We find that loss of tumorous traits is accompaniedby partial loss ofT-DNA. However, the postmeiotic tissues stillretain detectable T-DNA sequences.

MATERIALS AND METHODSMaterials. The crown gall tumor BT37 was incited on Ni-

cotiana tabacum var. Havana by A. tumefaciens strain T37,which carries the T37 Ti plasmid (35). A. Braun and R. Turgeonprovided the tumor tissue that had been cloned and then main-tained in culture for several years. Normal N. tabacum var.Xanthi callus and revertant tissues were grown on Murashigeand Skoog medium (38) supplemented with kinetin at 0.5 mg/liter and 1-naphthaleneacetic acid at 0.5 mg/liter. Normal N.tabacum var. Havana plants were grown from seeds providedby R. Turgeon. Bst I restriction endonuclease was supplied byRichard Meagher.

Plants from Tumor Tissue. BT37 teratoma tissue, which hadbeen maintained on Murashige and Skoog medium (38) withoutphytohormones, was transferred to the same medium supple-mented with kinetin at 1 mg/liter to induce the developmentof shoots and leaves (A. C. Braun and R. Turgeon, personalcommunications; ref. 33). Normal-appearing shoots and leaveswere observed 7-10 days after transfer. After these shoots hadreached 2-3 cm in height, they were placed on Murashige andSkoog medium without phytohormones. After 2 weeks, rootformation occurred. When rooted plants reached 5-8 cm, theywere transferred to soil. Initially, a plastic bag was placed overthe plant, and the pot was placed in a dish or water to maintainthe humidity. The plants were fed with Murashige and Skoogmedium without sucrose or-hormones. After growth to about.

Abbreviation: T-DNA, Ti plasmid DNA incorporated into crown galltumor.* This material was prepared in partial fulfillment of the requirementsof the Graduate School of the University ofWashington for the Ph. D.degree of F. Y.

t Present address: Laboratory ofMolecular Genetics, National Instituteof Neurological and Communicative Disorders and Stroke, NationalInstitutes of Health, Bethesda, MD 20205.

4151

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

4152 Biochemistry: Yang and Simpson

tA ;.:i .... : ....B

FIG. 1. Generation of plants from the BT37 tumor. (A) BT37 teratoma grown without phytohormones. (B) Normal-appearing shoots and leaves17 days after transfer of teratoma tissue to medium containing kinetin at 1 mg/liter. (C) Rooted plantlet obtained 21 days after transfer of normal-appearing shoots to medium without phytohormones.

10 cm, they were transferred to larger pots and cultivated in aplant growth room.

Detection of Plasmid DNA Sequences. Total DNA fromplant tissue, T37 Ti plasmid DNA from A. tumefaciens A208(39), and cloned fragments of the Ti plasmid from Escherichiacoli were isolated as described (12). Thomashow et al. (12) alsodetail the Southern (40) nitrocellulose transfer and hybridiza-tion procedures used in this study. Fragments from a partial orcomplete digest ofT37 Ti plasmid by BstEII were cloned in thevehicle pMB9 according to 1978 National Institutes of Healthguidelines.

RESULTSRegeneration of Plants from a Crown Gall Teratoma by

Manipulation of the Hormone Balance. When BT37, a clonedteratoma tissue (Fig. 1A), was placed on Murashige and Skoogmedium (38) containing kinetin (1 mg/liter), normal-appearingshoots grew out of the teratoma (Fig. 1B). When these shootswere transferred to a hormone-free medium, they developedroots (Fig. 1C). After transfer to soil, they formed mature plants.Out ofa total of 19 regenerated plants, 10 appeared completelynormal, 4 were considered slightly abnormal, having nonuni-formly pigmented leaves, and 5 were abnormal with severelydeformed leaves (Fig. 2). In some cases, individual shoots onthe same plant displayed different morphologies. The abnor-mal-appearing plants flowered but set fewer seed than thosewith a normal morphology.The first property of revertant plants tested was the ability

FIG. 2. Leaves from mature revertants. (A) Leaves from a plant

with normal morphology. (B) Abnormally pigmented leaves from aplant with slightly abnormal morphology. (C) Leaves from aplant withabnormal morphology.

FIG. 3. Hybridization of whole T37 Ti plasmid DNA with a South-ern blot of revertants, parental BT37, and control DNAs digested withBst I. Lanes a-h represent a 2-week exposure, and lanes a' and b' rep-resent a 2-day exposure of a and b. Each lane contains 7 ,ug of the fol-lowingDNA: lane a, one-copy reconstruction mixture (DNA equivalentto one copy of Ti plasmid DNA per diploid plant genome-i.e., 140 pgof T37 Ti plasmidDNA plus 7 ,ug of salmon sperm DNA); lane b, BT37teratoma; lane c, normal N. tabacum var. Havana leaf; lane d, normalN. tabacum var. Xanthi callus; lane e, normal-appearing leaves frommature reverted plant no. 10; lane f, abnormal-appearing leaves frommature reverted plant no. 1; lane g, normal-appearing young revertedplant; lane h, abnormal-appearing young reverted plant. The numbersat the left refer to the Bst I fragments of T37 Ti plasmid discussed inthe text. On the right are the molecular weights (x 10-6) of the in-dicated fragments.

afb aa*

w"a|

14a_

20

w7.3

r4.6

.-3.9

_.^....*

Proc. Natl. Acad. Sci. USA 78 (1981)

Proc. Natl. Acad. Sci. USA 78 (1981) 4153

Bst I | 20

BstEII1217b 149 11 14a |

63a $

I 18 121aL

T-DNA

Common DNA

FIG. 4. Restriction endonuclease map of the T-DNA region of T37 Ti plasmid (A. L. Montoya, F.-M. Yang, and M.-D. Chilton, personal com-munication). The T-DNA region in BT37 tumor is indicated (11, 13, 17). The common DNA sequences are highly conserved among diverse typesof Ti plasmids (24, 25).

ofthe tissue to grow in culture in the absence ofplant hormones.A leaf from each of the regenerated plants was cut in half andplaced on media with and without phytohormones. The rever-tant tissues grew only on media containing phytohormones. Thesecond property examined was the ability to produce nopaline(41). In contrast to the abundant production of nopaline by theparental BT37 tumor, nopaline was not detected in any rever-tant tissues. The third property investigated was the resistanceto reinfection by Agrobacterium. Strain A208 (39) containingthe same plasmid as that causing the parental tumor formedteratomatous tumors on revertant plants, and an octopine-typestrain (A277) formed unorganized tumors on revertant plants.This is in contrast to suppressed tumor cells, which displayedresistance to reinfection (36). Therefore, by all three propertiesexamined the revertant plants appeared normal and displayedno tumorous characteristics.

Detection ofPlasmid DNA Sequences in the Reverted PlantTissue. Another characteristic of crown gall tumor cells is thepresence of bacterial DNA sequences derived from the A. tu-mefaciens Ti plasmid. Therefore, we performed Southern blotexperiments (40), using DNA isolated from morphologicallydifferent types of reverted plant tissues. DNAs isolated fromseveral normal- and abnormal-appearing revertant plants weredigested with the restriction endonuclease Bst I, fractionatedby electrophoresis on agarose gels, transferred to nitrocellulosesheets, and then hybridized with 32P-labeled T37 Ti plasmidDNA.

Radioautography revealed the presence of some sequenceshomologous to plasmid DNA (Fig. 3, lanes e, f, g, and h). Ineach of the four DNA preparations, a total of four fragments,with molecular weights of 7.3 x 106, 4.6 X 106, 3.9 X 106, and2.1 X 106, were present that contained DNA sequences ho-mologous to pTi-T37 plasmid DNA. The fragment with a mo-lecular weight of 3.9 x 106 was also detected in DNA isolatedboth from normal tobacco callus and normal tobacco leaves (Fig.3, lanes c and d). The other three fragments detected in therevertants were not found in normal tobacco callus or leaves.The two internal plasmid Bst I fragments, 9 and 14a, presentin the parental BT37 tumor tissue (Fig. 3, lane b') were notdetected in any of the reverted plant tissues examined.A more sensitive experiment used a mixture of specific plas-

mid fragments as probes. These plasmid fragments, cloned inthe vehicle pMB9, contain the sequences between BstEII frag-ment 21a and BstEII fragment 21b (Fig. 4) and appear to includeall the T-DNA sequences found in the parental BT37 teratomatissue (compare lane b', Fig. 3 and lanes b and b', Fig. 5). Theincreased sensitivity reveals a small fourth internal Bst I frag-ment, 23a, not seen in Fig. 3, but present in the parental tissueand seen in other experiments using the entire Ti plasmid asprobe. Like the other internal fragments, Bst I fragment 23ais not present in the reverted tissues. The arrows in Fig. 5 in-dicate fragments with sequences homologous to the probe andhaving molecular weights of 8.3 x 106, 7.3 X 106, 4.6 X 106,and 2.1 x 106. The three smaller fragments are the same as

those detected with whole Ti plasmid as probe (Fig. 5). This,combined with the absence ofsequences homologous to T-DNAin normal tobacco callus (Fig. 5, lane h), conclusively shows thatthese three fragments, retained by the revertants, contain T-DNA.

Because the fragment with a molecular weight of 3.9 X 106was not observed in this hybridization experiment, we concludethat the plasmid sequences homologous to the normal plantDNA do not reside in the T-DNA region ofthe Ti plasmid. Thelargest fragment (8.3 x 106 daltons) is visualized with the T-DNA probe (Fig. 5) but not the entire Ti-plasmid probe (Fig.3). This fragment is not homologous with the cloning vehicle(data not shown) and may be a consequence ofeither incomplete

a b a b c d e f gh

3a6_ I9 initp

IWI:

14a _m*

20 :Iisr!

23a _;

8.3,- 7.3

w 4.6

w 2.1

FIG. 5. Hybridization of clone BstEH fragments 6, 9, 11, 14, 18,21a, and 21b with a Southern blot of parental BT37 teratoma, revert-ant, and control DNAs digested with Bst I. Lanes a-h represent an 8-day exposure, and lanes a' and b' represent a 17-hr exposure of a andb. Each lane contained 7 ug of the following DNAs: lane a, one-copyreconstruction mixture; lane b, BT37 teratoma; lane c, normal-ap-pearing leaves of mature reverted plant no. 11; lane d, abnormal-ap-pearing leaves of mature reverted plant no. 1; lane e, normal-appear-ing leaves of mature reverted plant no. 10; lane f, M1 plant from seedset by abnormal-appearing revertant no. 1; lane g, M1 plant from seedset by normal-appearing revertant no. 10; lane h, normal N. tabacumvar. Xanthi callus. The numbers at the left refer to theBst I fragmentsdiscussed in the text. On the right are the molecular weights (x 10-6)of the indicated fragments.

Biochemistry: Yang and Simpson

2311

6

1 9 1

4154 Biochemistry: Yang and Simpson

BAa b cde

Ca b d e c a b d e c

_.

1411

S.

20

-8 7.3-K7.3

3a4

....94-0

14a

20-

FIG. 6. Hybridization of different regions of the T-DNA with a Southern blot of revertant, parental BT37,and control DNAs digested withBstI. Each lane contains7 ,ug of the following DNAs: lanes a, one-copy recon-struction mixture; lanes b, BT37 teratoma; lanes c, nor-mal N. tabacum var. Xanthi leaf; lanes d, M1 plant fromseed set by abnormal-appearing revertant no. 1; lanes

-m4.6 e, normal-appearing leaves from mature reverted plantno. 10. In blotA the probe is pNW31C-8, 29-1 DNA (12),which contains the common DNA, the region of the oc-topine A6 Ti plasmid that is highly conserved amongdiverse types of Ti plasmids (24, 25). In blotB the probeis a clone containing BstEII fragments 9 and 14, whichrepresents the left end of the T-DNA. In blot C the probeis a clone containingBstEII fragments 18, 21a, and 21b,which represents the right end of the T-DNA. Becausethis probe includes BstEII fragment 21b, which lies out-

-w21 side the T-DNA on the left, Bst I fragment 20 is visiblein the one-copy reconstruction. The numbers at the leftof each blot refer to Bst I fragments of the T37 Ti plas-mid. To the right of each blot are the molecular weights(x 10-6) of the indicated fragments.

digestion with Bst I or insufficient sensitivity ofthe experimentwith the whole Ti plasmid as probe.We determined which portions of the T-DNA are present in

the revertant tissues by using individual probes that repre-sented different regions of the T-DNA. For example, no ho-mology was detected in the revertant tissue DNAs with a probethat contained the common DNA sequences (Fig. 6A). Further,the entire middle of the T-DNA has been eliminated, as evi-denced by the lack of homology to BstEII fragments 6 and 11(data not shown). On the other hand, when reverted plantDNAs were hybridized with probes that represent left and rightends of the T-DNA (BstEII fragments 9, 14, 18, and 21ab), allof the four previously detected fragments were identified (Fig.6B and C). Sequences homologous to the left end ofthe T-DNAwere found on three fragments (8.3, 7.3, and 2. 1 X 106 daltons)and sequences homologous to the right end ofthe T-DNA werefound on two fragments (4.6 and 2.1 X 106 daltons). Althoughthe precise quantity of T-DNA retained remains to be deter-mined, it is clear that only the ends of the T-DNA are left inthe reverted plant tissues.We examined the T-DNA content ofseveral revertants, some

with quite abnormal morphology and some that appeared nor-mal. The analysis included tissues from reverted young plantsas well as mature, seed-bearing revertants. As Fig. 3, 5, and 6show, they all retained the same T-DNA complement.

Analysis of F1 Plants Derived from Regenerated Plants. Inorder to study the progeny of the reverted plants and the fateof the plasmid DNA after meiosis, seeds arising from self-pol-lination were germinated. All of these Ml plants were normalin appearance regardless of whether the seeds were set by re-generated plants with normal or abnormal morphologies. Likethe parental revertant plants, none of the M1 plants exhibitedany tumorous properties.

Similarly, all of the homologous sequences detected in theparental revertant plants were still present in the M1 plants.When DNAs prepared from M1 plants were hybridized withcloned plasmid DNA fragments, hybridization patterns iden-tical to those of the parental plants were obtained (Fig. 5, lanesf and g; Fig. 6, lane e). These results indicate that all of theseforeign DNA sequences are retained even after the plant cellshave undergone meiosis.

DISCUSSION

We have demonstrated that a cloned crown gall tumor can formintact plants that appear to have normal physiology and func-tion. The loss of tumorous traits is accompanied by the loss ofmost of the incorporated bacterial plasmid DNA sequences.However, some ofthe T-DNA sequences remain stable throughmeiosis and thus are transmitted in the seed.

Analysis of the T-DNA indicates that the revertants lack thecentralT-DNA sequences found in the parental tissue but retainsequences homologous to the ends of the T-DNA. The se-quences that have been eliminated contain the common DNAsequences, a highly conserved region ofTi plasmids (24, 25) thatis incorporated into all tumors studied (12, 13, 17). Further,insertions and deletions within this region affect virulence (26-30). Thus, these sequences appear necessary for oncogenicityand tumor maintenance and their loss is probably directly re-lated to tumor reversal. Research on virally induced animal tu-mors has revealed a similar association between the loss of for-eign DNA (viral) and loss of tumorous traits (42-45).

Identical hybridization patterns were observed in DNApreparations from seven revertant plants that arose as shootsat different times and from different culture dishes. This sug-gests either that there is a common excision mechanism or thatthe seven revertants were derived from a common progenitorcell. We favor the latter because other revertants from the sametumor (discussed below) retain no detectable T-DNA. The Mlplants arising from seed set by the reverted plants retain thesame T-DNA sequences found in the parental plant. The sta-bility of the T-DNA in subsequent generations remains to bedetermined.

Manipulation of the hormone levels in the growth media re-sults in the production ofboth normal- and abnormal-appearingplants from the teratoma. The various plants are similar in thatthey have lost all of their tumorous characteristics and retainthe same T-DNA complement. Therefore, the differences inmorphology might be due to differences in chromosome com-plements occurring after the reversion event. The presence ofdifferent morphologies on individual shoots of the same revert-ant suggests that a revertant might arise from more than onecell (Armin C. Braun, personal communication). Whether the

Proc. Natl. Acad. Sci. USA 78 (1981)

--2,1

Proc. Natl. Acad. Sci. USA 78 (1981) 4155

revertants are chimeras remains to be determined. Regardlessof whether seeds were set by regenerated shoots with normalor abnormal morphology, all products ofself-fertilization appearnormal. Possibly the only products of meiosis are ones with arelatively normal karyotype.From the same cloned BT37 tumor used in these experi-

ments, Turgeon, Wood, and Braun obtained morphologicallynormal shoots by grafting the tumor to a healthy tobacco plant(35, 36). Tissue from various parts of the shoots retained malig-nant properties (35) and what appears to be at least one copyofthe entire T-DNA (11, 13). However, postmeiotic tissues thatappeared morphologically normal (36) contained no detectableT-DNA (11, 13). In contrast, our premeiotic revertant tissueretains neither malignant properties nor an entire copy of theT-DNA, and meiosis does not alter this residual T-DNA. Thus,we suggest that if meiosis acts either to cause or select for lossof foreign DNA, the loss is not due to the presence of foreignDNA per se.The reversion of crown gall tumors to normal functioning

plants with the retention ofsome ofthe foreign DNA sequencesthrough seed formation suggests that crown gall may be usefulin the genetic engineering ofplants. The results reported in thismanuscript indicate that the use of the Ti plasmid as a vectorneed not be restricted to vegetatively propagated crops.

We thank Dr. Armin Braun for the cloned parental BT37 teratomaline, Alice Montoya for assistance in the preparation of this manuscript,Michael Thomashow for criticisms of the manuscript, Sharon Bradleyfor typing the manuscript, Frank White for photographs ofplant tissues,and Mary-Dell Chilton for valuable discussions. We thank Dr. MiltonP. Gordon for performing the initial experiments that defined the con-ditions for the reversion ofBT37. The advice and encouragement of Dr.Eugene Nester are gratefully acknowledged. R. B. S. is a fellow in Can-cer Research supported by Grant DRG234F of the Damon Runyon-Walter Winchell Cancer Fund. This work was supported by NationalCancer Institute Grant CA 13015 and American Cancer Society GrantNP 336.

1. Zaenen, I., Van Larebeke, N., Teuchy, H., Van Montagu, M.& Schell, J. (1974) J. Mol Biol 86, 109-127.

2. Watson, B., Currier, T. C., Gordon, M. P., Chilton, M.-D. &Nester, E. W. (1975) J. Bacteriot 123, 255-264.

3. Van Larebeke, N., Engler, G., Holsters, M., Van den Elsaker,S., Zaenen, I., Schilperoort, R. A. & Schell, J. (1974) Nature(London) 252, 169-170.

4. Van Larebeke, N., Genetello, C., Schell, J., Schilperoort, R. A.,Hermans, A. K., Hernalsteens, J. P. & Van Montagu, M. (1975)Nature (London) 255, 742-743.

5. Braun, A. C. (1956) Cancer Res. 16, 53-56.6. Goldman, A., Thomas, D. W. & Morel, G. (1969) C. R. Hebd.

Seances Acad. Sci. Ser. D. 268, 852-854.7. Menage, A. & Morel, G. (1964) C. R. Hebd. Seances Acad. Sci.

Ser. D. 259, 4795-4796.8. Firmin, J. L. & Fenwick, R. G. (1978) Nature (London) 276, 842-

844.9. Chilton, M.-D., Drummond, M. H., Merlo, D. J., Sciaky, D.,

Montoya, A. L., Gordon, M. P. & Nester, E. W. (1977) Cell 11,263-271.

10. Merlo, D. J., Nutter, R. C., Montoya, A. L., Garfinkel, D. J.,Drummond, M. H., Chilton, M.-D., Gordon, M. P. & Nester,E. W. (1980) Mol Gen. Genet. 177, 637-643.

11. Yang, F.-M., Montoya, A. L., Merlo, D. J., Drummond, M. H.,Chilton, M.-D., Nester, E. W. & Gordon, M. P. (1980) MolGen. Genet. 177, 707-714.

12. Thomashow, M. F., Nutter, R., Montoya, A. L., Gordon, M. P.& Nester, E. W. (1980) Cell 19, 729-739.

13. Lemmers, M., De Beuckeleer, M., Holsters, M., Zambryski, P.,Depicker, A., Hernalsteens, J. P., Van Montagu, M. & Schell,J. (1980) J. Mot Biol. 144, 355-378.

14. Yadav, N. S., Postle, K., Saiki, R. K., Thomashow, M. F. & Chil-ton, M.-D. (1980) Nature (London) 287, 458-461.

15. Zambryski, P., Holsters, M., Kruger, K., Depicker, A., Schell,J., Van Montagu, M. & Goodman, H. M. (1980) Science 209,1385-1391.

16. Willmitzer, L., De Beuckeleer, M., Lemmers, M., Van Mon-tagu, M. & Schell, J. (1980) Nature (London) 287, 359-361.

17. Chilton, M.-D., Saiki, R. K., Yadav, N., Gordon, M. P. & Que-tier, F. (1980) Proc. Natl Acad. Sci. USA 77, 4060-4064.

18. Thomashow, M. F., Nutter, R., Postle, K., Chilton, M.-D.,Blattner, F. R., Powell, A., Gordon, M. P. & Nester, E. W.(1980) Proc. Natl, Acad. Sci. USA 77, 6448-6452.

19. Drummond, M. H., Gordon, M. P., Nester, E. W. & Chilton,M.-D. (1977) Nature (London) 269, 535-536.

20. Ledeboer, A. M. (1978) Dissertation (Univ. Leiden, Leiden, TheNetherlands).

21. Gurley, W. B., Kemp, J. D., Albert, M. J., Sutton, D. W. &Callis, J. (1979) Proc. Natl. Acad. Sci. USA 76, 2828-2832.

22. Yang, F.-M., McPherson, J. C., Gordon, M. P. & Nester, E. W.(1980) Biochem. Biophys. Res. Commun. 92, 1273-1277.

23. McPherson, J. C., Nester, E. W. & Gordon, M. P. (1980) Proc.NatL Acad. Sci. USA 77, 2666-2670.

24. Chilton, M.-D., Drummond, M. H., Merlo, D. J. & Sciaky, D.(1978) Nature (London) 275, 147-149.

25. Depicker, A., Van Montagu, M. & Schell, J. (1978) Nature (Lon-don) 275, 150-153.

26. Holsters, M., Silva, B., Genetello, C., Engler, G., Van Vliet, F.,deBlock, R., Villarroel, R., Van Montagu, M. & Schell, J. (1978)Plasmid 1, 456-467.

27. Koekman, B. P., Ooms, G., Klapwijk, P. M. & Schilperoort, R.A. (1979) Plasmid 2, 347-357.

28. Holsters, M., Silva, B., Van Vliet, F., Genetello, C., De Block,M., Dhaese, P., Depicker, A., Inze, D.,.Engler, G., Villarroel,R., Van Montagu, M. & Schell, J. (1980) Plasmid 3, 212-230.

29. Ooms, G., Klapwijk, P. M., Poulis, J. A. & Schilperoort, R. A.(1980)J. Bacteriol 144, 82-91.

30. Garfinkel, D. J. & Nester, E. W. (1980) J. Bacteriol 144, 732-743.

31. Lutz, A. (1966) C. R. Hebd. Seances Acad. Sci. Ser. X 262, 1856-1858.

32. Sacristan, M. D. & Melchers, G. (1977) Mot Gen. Genet. 152,111-117.

33. Einset, J. W. & Cheng, A. (1979) In Vitro 15, 703-708.34. Yang, F., Merlo, D. J., Gordon, M. P. & Nester, E. W. (1980)

Mot Gen. Genet 179, 223-226.35. Braun, A. C. & Wood, H. N. (1976) Proc. Natl Acad. Sci. USA

73, 496-500.36. Turgeon, R., Wood, H. N. & Braun, A. C. (1976) Proc. NatL

Acad. Sci. USA 73, 3562-3564.37. Yang, F., Montoya, A. L., Nester, E. W. & Gordon, M. P. (1980)

In Vitro 16, 87-92.38. Murashige, T. & Skoog, F. (1962) Physiol Plant. 15, 473-497.39. Sciaky, D., Montoya, A. L. & Chilton, M.-D. (1978) Plasmid 1,

238-253.40. Southern, E. M. (1975) J. MoL Biol 98, 503-517.41. Montoya, A. L., Chilton, M.-D., Gordon, M. P., Sciaky, D. &

Nester, E. W. (1977)J. Bacteriol 129, 101-107.42. Basilico, C., Gattoni, S., Zouzias, D. & Valle, G. D. (1979) Cell

17, 645-659.43. Steinberg, B., Pollock, R., Topp, W. & Botchan, M. (1978) Cell

13, 19-32.44. Frankel, A. E., Haapala, D. K., Neubauer, R. L. & Fischinger,

P. J. (1976) Science 191, 1264-1266.45. Eick, D., Stabel, S. & Doerfler, W. (1980)J. ViroL 36, 41-49.

Biochemistry: Yang and Simpson