Field performance and heavy metal concentrations of transgenic flue-cured tobacco expressing a mammalian metallothionein-P-glucuronidase gene fusion

J. E. B R A N D L E '

Agriculture Canada, Research Station, P. 0 . Box 186, Delhi, Ont., Canada N4B 2 W9

A N D

H. LABBE, J. HATTORI, A N D B. L. MIKI Agriculture Canada, Plant Research Centre, Ottawa, Ont., Canada KIA OC6

Corresponding Editor: R. J. Kemble

Received August 17, 1992

Accepted October 30, 1992

BRANDLE, J. E., LABBE, H., HATTORI, J., and MIKI, B. L. 1993. Field performance and heavy metal concentrations of transgenic flue-cured tobacco expressing a mammalian metallothionein-P-glucuronidase gene fusion. Genome, 36: 255-260.

Cadmium (Cd) is a nonessential heavy metal that can cause acute and chronic illness in humans. Some plant species such as tobacco (Nicotiana tabacum L.) tend to accumulate high levels of Cd in leaf tissue, the con- sumed portion of the plant. Tissue-specific expression of mammalian metallothionein has been suggested as a means of partitioning Cd in nonconsumed portions of transgenic plants. The purpose of the experiment reported here was to evaluate Cd concentration and agronomic performance of four field-grown transgenic tobacco lines har- bouring a metallothionein-P-glucuronidase (MG) gene fusion driven by the constitutive 35s promoter of cauli- flower mosaic virus. The trial was grown in a region of Canada known to have high background levels of Cd. The agronomic evaluation showed that some of the transgenic lines were equal to, while others performed more poorly than, the untransformed control for yield, days to I'lower, and leaf number. Gene expression measured by P-glucuronidase activity showed that all of the transgenic lines expressed the MG gene in the upper portion of the plant. One line did not express the MG gene in the roots. Cd levels in the leaf tissue of transformed lines were not significantly different from the untransformed control.

Key words: cadmium, P-glucuronidase, transformation, heavy metal tolerance.

BRANDLE, J. E., LABBE, H., HATTORI, J., et MIKI, B. L. 1993. Field performance and heavy metal concentrations of transgenic flue-cured tobacco expressing a mammalian metallothionein-P-glucuronidase gene fusion. Genome, 36 : 255-260.

Le cadmium (Cd) est un mktal lourd non-essentiel qui peut causer une maladie aigue et chronique chez les humains. Certaines espkces vkgktales, comme le tabac (Nicotiana tabacum L.), peuvent accumuler des concen- trations klevkes de Cd dans les tissus foliaires, lesquels font I'objet de consommation. L'expression spkcifique dans ces tissus d'une mktallothionkine mammalienne a kt6 suggkrke comme moyen de rkpartir le Cd dans les portions non consommkes de plantes transgkniques. L'expkrience prksentke ici a visk a kvaluer la concentration en Cd et la performance agronomique au champ de quatre lignkes de tabacs transgkniques porteuses d'une fusion gknique mktallothionkine-P-glucuronidase (MG), amorcke par la promoteur constitutif 35s du virus de la mosai'que du chou-fleur. L'expkrience a kt6 faite dans une rkgion du Canada connue comme ayant des teneurs klevkes en Cd. L'kvaluation agronomique, en fonction du rendement, du nombre de jours requis pour la florai- son et du nombre de feuilles, a montrk que certaines lignkes transgkniques ktaient kgales aux tkmoins non transformks, alors que d'autres ktaient infkrieures. L'expression gknique mesurke par I'activitk P-glucuronidase a rkvklk que toutes les lignkes transgkniques ont exprimk le gkne MG dans la partie supkrieure des plantes. Une lignke n'a pas exprimk le gkne MG dans les racines. La concentration de Cd dans le tissu foliaire des lignkes transformkes n'ont pas diffkrk significativement de celles des tkmoins non transformks.

Mots cle's : cadmium, P-glucuronidase, transformation, tolkrance aux mktaux lourds. [Traduit par la rkdaction]

Introduction Since the leaf tissue is the consumed portion of the

Cadmium (Cd) is a contaminant in a number of agri- cultural soils that results from either excessive phosphate fertilization (Williams 1977), use of sewage sludge as a soil amendment (Alloway et al. 1990), or naturally high background levels. Most crop species tend to accumulate Cd at the highest concentrations in the root tissue, fol- lowed by leaves, then by seeds or storage organs (Davis 1984). In contrast, some species, such as cabbage (Brassica oleracea L.), lettuce (Latuca sativa L.), and tobacco (Nicotiana tabacum L.), will accumulate high concentrations of Cd in leaf tissue rather than in roots.

' ~ u t h o r to whom all correspondence should be addressed.

plant, Cd can move into the human population, risking chronic exposure and threatening human health (Ryan e t al. 1982).

In animals, Cd is stably sequestrated by a class of pro- teins, the metallothioneins, which selectively bind heavy metals through thiolate bonds to cysteine residues (Nielson et al. 1985). Metallothioneins are not generally found in plants; however, cDNA corresponding to puta- tive metallothionein-like proteins have been cloned from pea, soybean, and Mimulus guttatus (Evans et al. 1990; Kawashima et al. 199 1; d e Miranda e t al. 1990). Plants normally respond to excess heavy metals by producing another class of small metal-binding peptides, the phy- tochelatins, which unlike metallothionein are not coded

Printed in Canada I I m p r ~ m e au Canada

Gen

ome

Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

NE

W M

EX

ICO

on

11/2

7/14

For

pers

onal

use

onl

y.

GENOME, VOL. 36, 1993

for by genes (Grill et al. 1987). The use of the Chinese hamster metallothionein I1 gene (CHMTII) to increase the capacity of plants to bind and detoxify Cd was first proposed by Lefebvre et al. (1987). Their experiments showed that the CHMTII gene expressed in Brassica campestris L. with a cauliflower mosaic virus vector dra- matically reduces the free Cd in transfected leaves.

Maiti et al. (1989) demonstrated that constitutive expression of a mouse metallothionein gene driven by the 35s promoter in transgenic burley tobacco (N. tabacum cv. Ky 14) yielded leaf Cd concentrations in 15-day-old seedlings that were up to 20% lower than that of untrans- formed controls. The authors concluded that Cd seques- tration in the roots of transgenic tobacco seedlings was responsible for reduced leaf Cd concentrations. Misra and Gedamu (1989) have shown that seedlings of both transgenic tobacco and oilseed rape (Brassica napus L.) expressed an introduced human metallothionein gene and showed greater tolerance to toxic levels of Cd. No field studies or evaluations of this system in adult plants have been reported. The purpose of this work was to assess the agronomic performance of field-grown transformed plants harbouring a P-glucuronidase-metallothionein (MG) gene fusion driven by the 35s promoter and to evaluate this system as a means of blocking Cd accumu- lation in' the leaf tissues of field-grown flue-cured tobacco. Other studies have shown that the fusion protein retained both Cd-binding capacity and P-glucuronidase (GUS) activity (J. Hattori et al., in preparation). Expres- sion of the hybrid protein under field conditions was mon- itored using the GUS reporter gene system (Jefferson et al. 1987). Concentration of Cd was measured in the root and leaf tissue of four field-grown transgenic, flue- cured tobacco lines and compared with an untransformed control.

Materials and methods The Agrobacterium tumefaciens strain GV3 101 harboured

pGV3850::p35S-MG. The vector p35S-MG (J. Hattori et a]., in preparation) harboured a 35s-nptll-nos gene (Sanders et al. 1987) as the selectable marker and a 35s-MTIIGUS-nos gene. The latter consisted of a NcoUPvuII fragment from the Chinese hamster MTII cDNA clone pCHMTII (Griffith et al. 1983) ligated in frame to the SmaIISstI fragment of pBI101, which contained the GUS gene (Jefferson et al. 1987).

Transgenic plants were generated by co-cultivation of leaf discs of N. tabacum cv. Delgold grown in vitro (Horsch et al. 1985). Selection on 100 mg1L kanamycin sulfate (Sigma Chemical Co.) was performed in the presence of 500 mglL cefotaxime (Roussel) on MS medium (Murashige and Skoog 1962) with B5 vitamins (Gamborg et a1 1968), 3% sucrose, 0.1 mglL NAA, and 1 mg/L BA. The tissues were transferred to media lacking growth regulators for root regeneration.

Segregation for quantitative GUS activity was calculated from the results of the screening tests described below. Seeds resulting from the self-pollination of four primary transfor- mants, DMGI, DMG2, DMG3, DMG4, and the untrans- formed control cv. Delgold were germinated on moist filter paper in Petri plates. Approximately 500 seedlings were removed from the Petri plates and transferred to 28 X 54 cm plug trays filled with soil-less potting mix. Following trans- fer, seedlings were grown in the greenhouse for 4 weeks and then each was screened qualitatively for GUS activity as described by Jefferson (1987), except a 24-h incubation at 37°C was used. Following incubation the individual tubes

Gen

ome

Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

NE

W M

EX

ICO

on

11/2

7/14

For

pers

onal

use

onl

y.

BRANDLE ET AL

Table I. Segregation analysis of four transgenic lines for kanamycin resistance (Kmr:Km') and GUS (GUS+:GUS-) activity

Line

DMG 1 49:20 0.58 0.3-0.5 193:70 0.37 0.3-0.5 DMG2 57: 18 0.04 0.7-0.9 188:52 1.42 0.2-0.3 DMG3 57: 16 0.37 0.3-0.5 177:63 0.20 0.5-0.7 DMG4 26: 12 0.88 0.3-0.5 193:47 3.70 0.05-0. I0

Ratios were tested for goodness of fit to a 3: 1 ratio. 'Represents the probability that the obserbed ratio fits the proposed ratio of 3: 1 .

TABLE 2. Quantitative GUS activity of greenhouse and field-grown transgenic lines (DMGI, DMG2, DMG3, and DMG4) and an untransformed control

(CV. Delgold)

nM MU.min ' .mgp ' protein

Conditions Organ DMGl DMG2 DMG3 DMG4 cv. Delgold

Field Leaf 23.6ah 6.2h 35.5uh 38.711 0.41. Greenhouse Leaf 17.2 10.0" 23.7 41.3 0.04" Greenhouse Stem 10.4 8.5* 12.4 12.1 0.2 Greenhouse Root 16.7 0.5 21.3 27.8 0.4*

N o ~ t : Means followed by the same letter were not significantly different at P = 0.05 *Represents a single observation.

containing the assay mixture were exposed to long wave ultraviolet light and scored as either GUS positive (GUS+) or GUS negative (GUS-). Segregation for kanamycin resistance was determined as described in Brandle and Miki (1992).

Following the qualitative GUS assay, seedlings corresponding to the GUS- tubes were discarded and those that were GUS+ were placed back in the greenhouse to be used for transplanting to the field. The seedlings from the untransformed control (CV. Delgold) were treated in the same manner as the trans- formants but were not screened for GUS activity. On June 8, 1989, all seedlings were transplanted to the field. Application to field test transgenic plants was made to the Plant Health Directorate, Food Production and Inspection Branch of Agriculture Canada, and acceptance was received prior to the execution of the field trial. The trial was planted as a ran- domized complete block design with four replications. Plots consisted of single rows containing 18 plants with 61 cm between plants. Rows were spaced 107 cm apart and the planting density was 15 377 plantsfha. Production practices were those normally employed for flue-cured tobacco. Dates of flowering, plant height, leaf number, leaf area of the top three leaves, and ground sucker weights were recorded. Harvest began on August 13 and approximately three to four mature leaves were harvested every 4-7 days until September 7. There were a total of five harvest dates. Leaf material was cured using standard practices for flue-cured tobacco. Follow- ing curing, the tobacco was weighed and graded, and yield and grade index were calculated.

Quantitative GUS activity was measured on five randomly selected plants from each plot. On July 15, a 5-cm disc was sampled from a leaf 15 cm long from each selected plant. The leaf discs were quick-frozen at -80°C and stored in a biofreezer (-80°C) for later analysis. The method of Jefferson (1987) was used for all GUS analyses. A small amount (30-60 mg fresh weight) of tissue was sub-sampled from the frozen leaf disc, weighed, and placed in an aliquot of lysis buffer. Tissue was broken up using a glass rod and then homogenized using a hand-held homogenizer (OMNI 1000, Omni Int'l). The fluorometric reaction was carried out in 1 mM 4-methyl-umbelliferyl P-D-glucuronide (4-MUG, Sigma

Chemical Co.) and terminated after 60 min by the addition of 1.9 mL of 0.2 M Na2C0,. Fluorescence was then measured using a spectrotluorimeter (model TK0100, Hoefer) with exci- tation at 365 nm and emission at 455 nm. The GUS activity measurements were normalized to total protein content as determined by the technique of Bradford (1976). All GUS activity measurements from field-grown materials were trans- formed, log (x + I ) , prior to the applicalion of the analysis of variance and means separation procedures.

Following the completion of leaf harvest, five plants were chosen at random from each plot, roots were sampled, bulked, and stored at -20°C for analysis of metal concentra- tion. A random sample of cured leaf tissue corresponding to each harvest date of each plot was also taken for analysis of metal concentration. Root and cured leaf tissue samples were lyophilized, ground (Wiley mill), passed through a 20 mesh screen and a weighed sample (1.00 g) digested in 10 mL con- centrated HNO,. Samples were refluxed, filtered, and brought up to I00 mL volume with distilled water. The digest was then analyzed for Cd concentration using a flame atomic absorption spectrophotometer (Spectra 30, Varian). Total alkaloid concentration i n the cured leaf tissue was determined on a weighted composite sample representing all five har- vests of each plot and using the method of Harvey et al. ( 1969) (model I1 Autoanalyzer, Technicon).

Results and discussion Genetic analysis of the four transgenic lines indicated

that both GUS activity and kanamycin resistance were inherited as a single, dominant Mendelian characteristic (Table I ) . Five GUS+ plants from each transgenic line were grown to flowering in the greenhouse, selfed, and seed collected. Kanamycin resistance of those offspring was evaluated as described earlier. All GUS+ plants were either homozygous for, or segregated for, kanamycin resistance (data not shown), indicating that GUS activity and kanamycin resistance co-segregate and are almost certainly linked.

Gen

ome

Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

NE

W M

EX

ICO

on

11/2

7/14

For

pers

onal

use

onl

y.

GENOME, VOL. 36, 1993

TABLE 3. Agronomic performance of four field-grown transgenic lines (DMG 1 , DMG2, DMG3, and DMG4) and an untransformed control (cv. Delgold)

Yield Grade Days to Leaf Total index weight Suckers flower Leaves Height areaA alkaloids

(kglha) ($/kg) (kglha) (days) (no.)* (cm) (cm-)' (mglg)

DMGl 1909bc 422ab 5.4hc 57.8bc 14.4h 90.7u 81 la 2 5 . 6 ~ DMG2 2046ab 418b 9.6b 59.8ah 14.3h 9 4 . 2 ~ 777u 26.8u DMG3 1963hc 424a 8.3bc 5 6 . 5 ~ 14.7h 91.5u 738n 2 7 . 0 ~ DMG4 1 8 3 7 ~ 425a 1 9 . 2 ~ 58.0hc 14.3h 90.5u 685a 2 9 . 9 ~ Delgold 2160a 425a 1 . 9 ~ 6 0 . 3 ~ 1 7 . 0 ~ 8 8 . 1 ~ 763n 2 5 . 2 ~ SE 63 2 0.8 0.6 0.5 2.6 6 4 2.3

NOTE: Means followed by the same letter were not significantly different at P = 0.05. *Refers to the number of leaves remaining after topping. ' ~ e f e r s to the area of the top three leaves at full expansion.

TABLE 4. Concentration of Cd (mglkg) in leaf and root tissue of transformed (DMGI, DMG2, DMG3, and DMG4) and untrans- formed (cv. Delgold) tobacco lines grown at the Delhi Research

Station during 1990

Leaf*

Line Root I 2 3 4 5

DMG 1 0.6 4.8 3.1 2.7 2.4 2.2 DMG2 0.5 4.9 3.1 2.8 2.5 2.2 DMG3 0.6 4.8 3.1 2.8 2.3 2.2 DMG4 0.7 4.5 3.2 2.7 2.4 2.2 cv. Delgold 0.5 4.3 3.1 2.8 2.4 2.3 SE 0.1 0.1 0.1 0.1 0.1 0.2

*The numbers 1-5 represent the harvest date and leaf position, starting harvest from the bottom ( I ) and ending at the top ( 5 ) of the plant.

Histochemical staining of GUS activity was performed on 2- to 3-week-old seedlings grown in vitro. As shown in Fig. 1 GUS staining was apparent in all major vegetative organs of DMG 1 , DMG3, and DMG4; however, consider- able variation in the intensity of staining existed among the roots. This was also observed among more mature seedlings. Cross sections of leaf, stem, and roots revealed staining throughout the tissues including root hairs (data not shown). High levels of GUS activity were also measured in these organs using the fluorometric assay (Table 2). The data are consistent with the pattern of expression conferred by the 3 5 s promoter (Benfey et al. 1989). The transgenic line DMG2 differed from the others in that the staining of the stems and roots was very weak, whereas the staining of the leaves was similar to DMGl , DMG3, and DMG4 (Fig. 1). Fluorometric analysis of GUS activity confirmed the absence of GUS activity in the roots (Table 2). The repression of GUS activity in the roots of DMG2 is unexplained and may relate to position effects often encountered with the 3 5 s promoter (Benfey et al. 1989).

The GUS component of the MG fusion protein pro- vided a valuable marker for monitoring the presence of MG in the tissues of field-grown plants. All of the trans- genic lines showed GUS activity in the field at levels similar to those of greenhouse plants; however, there were differences in levels of expression among lines (Table 2). DMGI, DMG3, and DMG4 had similar levels of GUS activity in leaves, while DMG2 was significantly lower.

GUS activity in the untransformed check was negligible. Results of the field experiments showed that the agro-

nomic performance of the transgenic lines was, for some characters, equal to that of the untransformed control. However, some lines were either lower yielding, had lower grade index, or were earlier flowering than the untransformed control (Table 3). Sucker weights were significantly higher and leaf numbers were lower in the transgenic lines than in the untransformed controls.

Reduced yield and grade index have a negative impact on economic returns and excessive ground sucker growth interferes with the harvesting process. There is a strong positive correlation between flowering date and leaf num- ber (White et al. 1979), and flowering date and yield. Therefore, the low leaf numbers and reduced yields observed in the transgenics may have been at least in part due to early flowering. Height, area of the three top leaves, and total alkaloid concentration of transformed and untransformed lines were not significantly different.

The exact reasons for phenotypic differences among the transgenic lines are not clear and it would seem unlikely that the presence of the NPTII, CHMTII, or the GUS genes could have a direct effect on plant perfor- mance, especially given the presence of transgenic lines whose performance for some characteristics is equal to the untransformed control. However, Lefebvre (1990) has reported that glucosinolate biosynthesis was suppressed in B. carnpestris plants harbouring a metallothionein gene. It is, therefore, possible that the MG gene fusion indirectly affected biochemical processes related to nor- mal growth and development. Similar variation in the performance of transgenic lines has been reported with other constructs in tobacco (Brandle and Miki 1992) and tomato (Nelson et al. 1988). The source of the variation is probably the tissue culture step in the transformation process. Genetic variation among plants regenerated from tissue culture is common (Skirvin 1978). Therefore, regeneration of plantlets from leaf discs probably resulted in the introduction of somaclonal variation that affected the agronomic performance of the transgenic lines. Selection among transgenic lines or backcrossing will be required to completely recover the phenotype of the original parent.

Cd concentrations in the root and leaf tissues of trans- formed and untransformed lines were not significantly different (Table 4). These results indicate that the MG

Gen

ome

Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

NE

W M

EX

ICO

on

11/2

7/14

For

pers

onal

use

onl

y.

BRANDLE ET A L . 259

protein was not immobilizing Cd in the roots to a degree that would result in the exclusion of Cd from the leaf tis- sues. There did, however, appear to be a trend towards higher Cd levels in the roots of D M G l , DMG3, and DMG4, and the lower leaves (harvest 1) of all of the trans- genic lines, indicating that some binding could be taking place in the lower part of the plant. The absence of MG expression in the roots of DMG2 may correlate with the control levels of Cd in the roots of these plants and higher than control levels observed in the lower leaves. The concentrations of Cd in the upper part (harvests 2-4) of the transgenic lines were similar to that of the untrans- formed control, indicating that Cd was moving to the upper part of the plant unencumbered.

A number of explanations for these results are possi- ble. The concentration of available Cd in the soil solu- tion resulted in plant uptake rates that were in excess of the binding capacity of the MG produced by the plant. Levels of MG expression may not have been sufficient to compete with other, more mobile, Cd-binding molecules. Alternatively, expression of the MG gene may not have been sufficient in cell types involved in Cd uptake and translocation. The hybrid nature of the protein may have also negatively affected the ability of the MG protein to bind Cd, although bacterial expression studies d o not indicate that this is the case (J. Hattori et al., in prepara- tion). It is also possible that the Cd-MG complex is unstable, which allows greater mobility of Cd within the plant than would be expected given the normal multiyear half-life of metallothionein-Cd complexes in humans (Chung et al. 1986).

Leaf Cd concentrations were highest in the lower leaves and decreased towards the upper portion of the plant, in a manner similar to that reported by Wagner and Yeargan (1986) for burley tobacco and that reported by Frank et al. (1977) for flue-cured tobacco. This pattern was the same for transformed and untransformed lines. The accumulation was probably not age dependant as hypothesized by Wagner and Yeargan (1986), because in flue-cured tobacco all leaves are approximately the same age when harvested. The decrease in Cd concentration towards the top of the plant is more likely the result of binding in the lower part of the plant, either as a result of phytochelatins or other metal binding proteins, leading to reduced translocation to the upper leaves.

The results of this experiment indicate that expression of the MG gene does not significantly alter patterns of Cd accumulation in flue-cured tobacco. The exact rea- sons for the failure of the MG gene to block Cd accumu- lation in leaf tissues are unclear and further studies will be required to clarify the issue.

Alloway, B.J., Jackson, A.P., and Morgan, H. 1990. The accumulation of cadmium by vegetables grown on soils contaminated from a variety of sources. Sci. Total Environ. 91: 223-236.

Benfey, P.N., Ren, L., and Chua, N.H. 1989. The CaMV 35s enhancer contains at least two domains which can confer different developmental and tissue specifications. EMBO J. 8: 2 1 95-2202.

Bradford, M.M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein dye binding. Anal. Biochem. 72: 248-254.

Brandle, J.E., and Miki, B.L. 1992. Agronomic performance of sulfonylurea resistant transgenic flue-cured tobacco lines grown under field conditions. Crop Sci. In press.

Chung, J., Nartey, N.O., and Cherian, M.G. 1986. Metal- lothionein levels in liver and kidney of Canadians - a potential indicator of environmental exposure to cadmium. Arch. Environ Health, 41: 3 19-323.

Davis, R.D. 1984. Cadmium - a complex environmental problem. 11. Cadmium in sludges used as fertilizer. Experientia, 40: 1 17-234.

de Miranda, J.R., Thomas, M.A., Thurman, D.A., and Tomsett, A.B. 1990. Metallothionein genes from the flow- ering plant Mimulus gutrutus. FEBS Lett. 260: 277-280.

Evans, I.M., Gatehouse, L.N., Gatehouse, J.A., Robinson, N.J., and Croy, R.R.D. 1990. A gene from pea (Pisum scttivum L.) with homology to metallothionein genes. FEBS Lett. 262: 29-32.

Frank, R., Braun, H.E., Holdrinet, M., Stonefield, K.I., Elliot, J.M., Zilkey, B.F., Vickery, L., and Cheng, H.H. 1977. Metal contents and insecticide residues in tobacco soils and cured tobacco leaves collected in southern Ontario. Tob. Sci. 31: 40-45.

Gamborg, O.L., Miller, R.A., and Ojima, C. 1968. Nutrient requirements of suspension cultures of soybean root cells. Exp. Cell Res. 50: 151-158.

Griffith, B.B., Walters. R.A., Enger, M.D., Hildebrand, C.E., and Griffith, J.K. 1983. cDNA cloning and nucleotide sequence comparison of Chinese hamster metallothionein I and 11. Nucleic Acids Res. 11: 901-910.

Grill, E., Winnacker, E.L., and Zenk, M.H. 1987. Phyto- chelatins, a class of heavy-metal-binding peptides from plants, are functionally analogous to metallothioneins. Proc. Natl. Acad. Sci. U.S.A. 84: 439-443.

Harvey, W.R., Stahr, H.M., and Smith, W.C. 1969. Automated determination of reducing sugars and nicotine alkaloids on the same extract of tobacco leaf. Tob. Sci. 13: 13-1 5.

Horsch, R.B., Fry, J.E., Hoffman, N.L., Eichholtz, D., Rogers, S.G., and Fraley, R.T. 1985. A simple and general method of transferring genes into plants. Science (Washington, D.C.), 227: 1229- 123 1 .

Jefferson, R.A. 1987. Assaying chimeric genes in plants: the GUS gene fusion system. Plant Mol. Biol. Rep. 5: 387-405.

Jefferson, R.A., Kavanagh, T.A., and Bevan, M.W. 1987. GUS fusions: beta-glucuronidase as a sensitive and versa- tile gene fusion marker in higher plants. EMBO J. 6: 390 1-3907.

Kawashima, I., Inokuchi, Y., Chino, M., Kimura, M., and Shimuze, N. 1991. Isolation of a gene for a metallo- thionein-like protein from soybean. Plant Cell Physiol. 32: 9 13-926.

Lefebvre, D.D. 1990. Expression of mammalian metal- lothionein suppresses glucosinolate synthesis in Brussica campestris. Plant Physiol. 93: 522-524.

Lefebvre, D.D., Miki, B.L., and Laliberte, J. 1987. Mammalian metallothionein function in plants. Biotech- nology, 5: 1053- 1056.

Maiti, I.B., Wagner, G.J., Yeargan, R., and Hunt, A.G. 1989. Inheritance and expression of the mouse metallothionein gene in tobacco. Plant Physiol. 91: 1020-1024.

Misra, S., and Gedamu, L. 1989. Heavy metal tolerant trans- genic Brassica napus L. and ~ i c o t i a n a tabacum L. plants. Theor. Appl. Genet. 78: 16 1 - 168.

Murashige, T., and Skoog, F. 1962. A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol. Plant 15: 473-497.

Nelson, R.S., McCormick, S.M., Delannay, X., Dube, R., Layton, J., Anderson, E.J., Kaniewska, M., Proksch, R.K., Horsch, R.B., Rogers, S.G., Fraley, R.T., and Beachy, R.N. 1988. Virus tolerance, plant growth, and field performance

Gen

ome

Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

NE

W M

EX

ICO

on

11/2

7/14

For

pers

onal

use

onl

y.

260 GENOME. VOL. 36. 1993

of transgenic tomato plants expressing coat protein from tobacco mosaic virus. Biotechnology, 6: 403-409.

Nielson, K.B., Atkin, C.L., and Winge, D.R. 1985. Metal- binding configurations in metallothionein. Biol. Chem. 260: 5342-5350.

Ryan, J.A., Pahren, H.R., and Lucas, J.B. 1982. Controlling cadmium in the human food chain: a review and rationale based on health effects. Environ. Res. 28: 25 1-302.

Sanders, P.R., Winter, J.A., Barnasm, A.R., and Rogers, S.A. 1987. Comparison of cauliflower mosaic virus 35s and nopaline synthase promoters in transgenic plants. Nucleic Acids Res. 15: 1543-1 558.

Skirvin, R.M. 1978. Natural and induced variation in tissue culture. Euphytica, 27: 24 1-266.

Wagner, G.J., and Yeargan, R. 1986. Variation in cadmium accumulation potential and tissue distribution of cadmium in tobacco. Plant Physiol. 82: 274-279.

White, F.H., Pandeya, R.S., and Dirks, V.A. 1979. Correlation studies among and between agronomic, chemical, physical and smoke characteristics in flue-cured tobacco (Nic-otiana tclhacum L.). Can. J . Plant Sci. 59: 1 1 1-1 20.

Williams, C.H. 1977. Trace metals and superphosphate: toxicity problems. J. Aus. Inst. Agric. Sci. 43: 99-109.

Gen

ome

Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

NE

W M

EX

ICO

on

11/2

7/14

For

pers

onal

use

onl

y.