|
|
||||||||
|
Plant Physiol, March 2001, Vol. 125, pp. 1216-1227 A Complete Set of Maize Individual Chromosome Additions to the Oat Genome1Plant Molecular Genetics Institute (R.G.K., O.R.-L., M.I.V., R.J.O., S.B.M., G.C., E.V.A., W.E.O., C.D.R., A.O.S., S.M.L., H.A.Z., R.L.P.) and Plant Science Research Unit, United States Department of Agriculture-Agricultural Research Service (USDA-ARS) (H.W.R.), Department of Agronomy and Plant Genetics, University of Minnesota, 411 Borlaug Hall, 1991 Buford Circle, St. Paul, Minnesota 55108-6026
All 10 chromosomes of maize (Zea mays, 2n = 2x = 20) were recovered as single additions to the haploid complement of oat (Avena sativa, 2n = 6x = 42) among F1 plants generated from crosses involving three different lines of maize to eight different lines of oat. In vitro rescue culture of more than 4,300 immature F1 embryos resulted in a germination frequency of 11% with recovery of 379 F1 plantlets (8.7%) of moderately vigorous growth. Some F1 plants were sectored with distinct chromosome constitutions among tillers of the same plant and also between root and shoot cells. Meiotic restitution facilitated development of un-reduced gametes in the F1. Self-pollination of these partially fertile F1 plants resulted in disomic additions (2n = 6x + 2 = 44) for maize chromosomes 1, 2, 3, 4, 6, 7, and 9. Maize chromosome 8 was recovered as a monosomic addition (2n = 6x + 1 = 43). Monosomic additions for maize chromosomes 5 and 10 to a haploid complement of oat (n = 3x + 1 = 22) were recovered several times among the F1 plants. Although partially fertile, these chromosome 5 and 10 addition plants have not yet transmitted the added maize chromosome to F2 offspring. We discuss the development and general utility of this set of oat-maize addition lines as a novel tool for maize genomics and genetics.
Sexual inter-species hybridization is a powerful tool for understanding genome structure and interaction in higher plants. Breaking through inter-species incompatibility also enables chromosome engineering with horizontal (species-to-species) transfer of characters to which plant researchers and breeders have no access when utilizing only intraspecies gene pools. The more remotely related the parental genomes, the more the gene pool may be enriched. Most of the favored host plants for chromosome engineering are
allopolyploids or amphidiploids. Although in rare cases pure diploid
species do tolerate alien chromosome additions, e.g. rye (Secale
cereale; Kynast, 1986 Oat (Avena sativa) and maize (Zea mays) are the
most remotely related plant species of which we are aware that can be
sexually hybridized and produce stable fertile partial hybrids. When
oat × maize fertilizations are accomplished, a primary
inter-species hybrid is generated that usually completely eliminates
the maize chromosomes in early phases of embryogenesis. Thus, the
resulting F1 hybrid becomes an allohaploid oat
plant (Rines and Dahleen, 1990 A novel situation can occur in oat × maize crosses in that
F1 hybrids may retain one or more maize
chromosome(s) added to the haploid oat complement (Rines et al., 1995 The addition of a maize chromosome to oat also enables analyses of the expression of homozygous maize allelic pairs in their native chromosomal environment in interaction with the oat host genome.
Primary Inter-Species F1 Hybrids All 10 maize chromosomes were recovered as single additions to the entire haploid complement of oats. These maize chromosome additions were among 379 F1 plants generated from crosses of three different lines of maize to about 60,000 emasculated florets of eight different lines of oat followed by in vitro embryo rescue culture. The time course for in vitro germination of the F1 embryos extended up to 9 weeks with an average of about 2 weeks. After the 9 weeks, embryos were considered dead and no longer maintained if there was no development of shoot and root. By this criterion, the total of 473 germinated F1 embryos account for a germination frequency of about 11% of more than 4,300 immature F1 embryos placed on culture medium. When the F1 embryos germinated and developed shoots of about 1 cm, plantlets were transferred from constant dark to short days (11 h, constant 20°C). The initial period in light was a very critical stage. Among plantlets that attained the size of 1-cm shoot length, about 20% (94 F1 embryos) suffered from absence of root development. At about 5-cm shoot growth, vigorous plants were transplanted into soil mix and grown as described in "Materials and Methods." In general, all F1 plantlets showed weak morphology and slow growth after germination. Morphological differences appeared among the F1 plants depending on the added maize chromosome and the oat genetic background. All F1 plants had an enhanced sensitivity to changing growth conditions, such as temperature and photoperiod, compared with normal oats. When a plant survived transplantation and started to grow after 1 or 2 weeks, 4- to 5-cm pieces of the leaf blade were clipped and genomic DNA extracted. A total of 379 F1 plants were tested for presence versus absence of maize chromosome(s). A total of 135 F1 plants possessed one or more maize chromosome(s) as indicated by a PCR product generated using primers specific for the maize sequence long terminal repeat (Grande 1-LTR; Fig. 1). Grande 1-LTR is a highly repetitive dispersed repeat present on every maize chromosome. To verify the number of maize chromosomes in maize-positive F1 plants, chromosomes in root tip cells were counted and genomic in situ hybridization (GISH) applied as described in "Materials and Methods." Two examples are shown in Figure 2. Maize chromosome identities were then determined by PCR of leaf tissue DNA using maize chromosome-specific simple sequence repeat (SSR) markers (Fig. 3). For each maize chromosome several primer sets were selected from the Maize Genome Database (http://www.agron.missouri.edu/) and tested for maize specificity against genomic DNA of the oat lines. Specificity was defined by presence of a PCR product in maize cv Seneca 60 with no detected product from DNA of oat or by PCR products in maize and in oat DNA, but with length polymorphisms in maize compared with oat as detected upon electrophoresis in an agarose gel. An additional criterion for suitability of the marker of interest was that the marker had to be present on both parental chromosomes of the maize Seneca 60 with the above described specificity against oat. Because Seneca 60 is an F1 hybrid, markers were tested also on a representative Seneca 60 F2 progeny set (data not shown). By these criteria only 10 markers (14% of the markers tested) were chosen as most suitable and reliably reproducible (Table I) under our standard SSR assay conditions. Another 25 markers produced maize chromosome-specific bands but were less reproducible or segregating in the Seneca 60 F2 test population. Based on the identification by SSRs, each of the 10 maize chromosomes showed retention in F1 plants both as single additions and in combination with other maize chromosomes (Fig. 4). Maize chromosome 5 was most frequently retained (44 times). Maize chromosome 3 and 10 were most rarely retained (seven times each). Some F1 plants were sectored with distinct chromosome constitutions between tillers of the same plant. The presence and number of maize chromosomes also differed between shoot tissues (PCR analysis) and root tips (GISH analysis) in some F1 plants.
The numbers of retained versus eliminated maize chromosomes among
various oat × maize F1 genotypes are
summarized in Table II, including
previous (Riera-Lizarazu et al., 1996
Our data involving multiple chromosome retention revealed no apparent preferential combination of specific maize chromosomes in the F1 plants. However, our data show also that chromosome 3 appears to be the most excluded chromosome from multiple combinations among surviving plants in the oat × maize backgrounds we analyzed. Chromosome 3 was observed only four times in combination with other maize chromosome(s). Maize Chromosome 1 Additions Maize chromosome 1 was observed 19 times among the F1 plants. The chromosome appeared five times as a single addition and 14 times in combination with other maize chromosomes (Fig. 4). Three lines are maintained: OMAm1.1 as a haploid monosomic chromosome 1 addition in the form of a vegetative tiller clone, OMAd1.7 as a fertile disomic chromosome 1 addition line, and OMAd1.12/d9.10 as a double disomic chromosome 1 and 9 addition line (Fig. 5). To identify each addition plant, their designation follows the nomenclature outlined in "Materials and Methods."
OMAm1.1 is tillered and semifertile. No positive maize chromosome 1 addition plant was found among more than 200 F2 offspring, indicating a lack of maize chromosome 1 transmission in the oat cv Preakness background. The F1 plant has been kept under short-day condition for more than 2 years by frequently breaking the plant into parts at its shoot base and repotting to generate fresh tillers (tiller clones). The clones are still growing with development of new tillers useful for DNA extraction. The cloned plants flower under either long day or short day and appear to have lost the long-day response common in oats. A severe response to long-day exposure is shorter leaf blades and pollen sterility. OMAm1.7 looks much like Starter oat but is a little shorter in height and has a tendency toward more erect leaves as in OMAm1.1 (Fig. 6A). Also, this chromosome 1 addition plant shows the same photoperiod response as the OMAm1.1 plant with highest vigor, flowering, and seed set under short-day conditions. A second remarkable character is its chimerism in maize chromosome transmission. The plant had 108 F2 seeds distributed over 11 panicles, with only three F2 plants testing positive for maize chromosome 1. These three plants had 44 chromosomes including a pair of maize chromosomes 1 (OMAd1.7). The three seeds were set on one panicle. All other panicles had offspring negative for maize chromosome 1 (105 seeds) or were sterile. There was no variation among leaves in the tillers with respect to presence versus absence of maize chromosome 1. The F1 plant is still being maintained as a tiller clone with the aim to produce more offspring for analysis of chimerism in plants. The maize-positive F2 plants were fertile and showed sectoring again when the F3 offspring were cytologically analyzed.
The third hybrid (OMAd1.12/d9.10) was a double monosomic addition for maize chromosomes 1 and 9. The plant looked like regular oat, but was shorter and exhibited accelerated senescence. The plant was fertile with 76 seeds: 10 seeds with both maize chromosomes transmitted, four seeds with no chromosome 1 and only one or two chromosome 9, and 51 seeds without a maize chromosome. Six seeds did not germinate. Maize Chromosome 2 Additions Maize chromosome 2 was observed 32 times among the F1 plants. The chromosome appeared 11 times as single additions and 21 times in combination with six of the other maize chromosomes (Fig. 4). Seven different fertile lines are established (OMAd2.1, OMAd2.2, OMAd2.3, OMAd2.4, OMAd2.8, OMAd2.10, and OMAd2.12). Although maize chromosome 2 single additions were found also in GAF Park and Sun II oats, all fertile addition lines are in Starter oat. All F1 plants were vigorous and showed good vegetative growth with high tillering capacity. Compared to haploid Starter, the positive maize chromosome 2 plants were later maturing and had an intense bluish leaf color (Fig. 6B). Stems were more heavily covered by wax than Starter oat. The disomic addition offspring maintained these features with remarkable consistency. Chromosome 2 addition lines have the highest fertility among all oat-maize addition lines produced to date. Maize Chromosome 3 Additions Maize chromosome 3 was observed seven times among the
F1 plants. Chromosomes 3 and 10 are the most
rarely retained maize chromosomes. Chromosome 3 appeared three times as
single additions and four times in combination with other maize
chromosomes (Fig. 4). The multiple additions and a single addition in
Starter oat died at the four-leaf stage. The remaining two lines were
fertile Maize Chromosome 4 Additions Maize chromosome 4 was observed 19 times among the F1 plants. The chromosome appeared six times as single additions and 13 times in combination with six of the other maize chromosomes (Fig. 4). One chromosome 4 was found in combination in OMAm4.9, which derived from the cross F1 (MN97201-1 × MN841801-1) × Seneca 60. A second chromosome 4 (OMAm4.15) was found in crosses with Kanota oat. All the other retentions occurred in crosses with Starter. Seven F1 plants were fertile and had transmission of the maize chromosome 4. The lines OMAm4.4 and OMAm4.5 descended from Starter × A188 crosses. All offspring with disomic chromosome 4 additions have light green leaves (Fig. 6B), some lighter than Starter oat. All addition plants were earlier maturing and had small seed. Fertility is moderate and independent of whether Seneca 60 or A188 was the chromosome 4 donor. Maize Chromosome 5 Additions Maize chromosome 5 was observed 44 times among the F1 plants. The chromosome appeared 14 times as single additions and 30 times in different combinations involving all other maize chromosomes (Fig. 4). Thus, chromosome 5 is the most frequently retained maize chromosome in oat. All chromosome 5-positive F1 plants tend to develop fewer but thicker tillers. Plants accelerate development after photoperiod induction and senesce earlier than the corresponding haploid oat genotypes. Two single maize chromosome 5 addition F1 plants are being maintained as tiller clones under short-day conditions. OMAm5.9 is an addition to haploid Starter oat. OMAm5.17 is an addition to a haploid F1 (MN97201-1 × MN841801-1) oat line. The most common striking character of these two lines is the development of branched shoots, though more pronounced in OMAm5.17 than in OMAm5.9 (Fig. 6D). The stalk branches were wrapped together in one leaf sheath, concordant with the heavy thick stems. When shifted into long day, OMAm5.9 and OMAm5.17 flowered and set seed. Their phenotypes were similar and of oat type. Moderate fertility provided for more than 160 F2 offspring. Absence of chromosome 5 in all offspring demonstrated a lack of female and male transmission for the maize chromosome 5 in different oat backgrounds tested to date. Maize Chromosome 6 Additions Maize chromosome 6 was observed 17 times among the
F1 plants. The chromosome appeared only four
times as single additions and 13 times in different combinations with
the other maize chromosomes (Fig. 4). Two single addition
F1 plants were fertile and produced disomic
addition offspring (OMAd6.1 and OMAd6.5), which are maintained as
established lines. Both lines descended from Starter × Seneca 60 crosses. Both F1 plants tillered in short-day
photoperiod. Younger plants were indistinguishable from haploid Starter
oat, but older plants developed necrotic and chlorotic spots on the leaf blades similar to disease lesion mimic mutants (Neuffer, 1994 Maize Chromosome 7 Additions Maize chromosome 7 was observed 31 times among the F1 plants. The chromosome appeared 10 times as single additions and 21 times in different combinations involving all other maize chromosomes (Fig. 4). Four of the single addition F1 plants were fertile; OMAm7.1 was derived from a GAF Park × Seneca 60 cross amd OMAm7.6, OMAm7.7, and OMAm7.10 were derived from Starter × Seneca 60 crosses. A common feature of these additions was a reduced stature. Leaf color and panicle shape were very similar to those of their oat parents (Fig. 6F). The cytogenetic behavior was distinct in comparison to other addition plants. Chromosome 7 transmission was distorted. Disomic addition plants produced 16% monosomic and about 12% nonaddition offspring. Through further generations the transmission frequencies seemed to increase, but plants still show a low level of chromosome instability. Maize Chromosome 8 Additions Maize chromosome 8 was observed 27 times among the F1 plants (Fig. 4). The chromosome appeared 10 times as single additions and 17 times in different combinations involving all other maize chromosomes except maize chromosome 3. All oat lines except MN97201 retained maize chromosome 8. Three positive F1 plants were fertile; however, only two plants transmitted the chromosome to the next generation. OMAm8.1 (GAF Park × Seneca 60) produced one positive F2 offspring with 42 chromosomes. Among its F3 offspring, one positive monosomic addition (2n = 6x + 1 = 43) appeared. The monosomic addition line originates from this genotype. The second fertile F1 plant (OMAm8.2) was derived from a Starter × bz1 mum-9 cross. It produced one maize chromosome 8-positive F2 offspring among 19 offspring plants analyzed. Three F2 seeds did not germinate. Although the maize chromosome identity has been shown by SSR marker application, the chromosomes have not yet been counted. This plant is only 5 weeks old but lacks the vigor of Starter oat. Maize Chromosome 9 Additions Maize chromosome 9 was observed 31 times among the F1 plants. The chromosome appeared 12 times as single additions and 19 times in different combinations with the other maize chromosomes (Fig. 4). Five different fertile F1 plants were derived from Starter × Seneca 60 (OMAm9.1, OMAm9.2, OMAm9.6, OMAm9.8, and OMAm9. 9). One fertile F1 plant was derived from Kanota × Seneca 60 (OMAm9.5). One fertile F1 plant was derived from GAF Park × Seneca 60 (OMAm9.17). Six F1 plants produced diploid offspring with disomic additions of chromosome 9. One F1 plant produced a diploid F2 plant with disomic addition of chromosome 9 plus disomic addition of chromosome 1 (OMAd1.12/d9.10). All disomic addition F2 plants inherited a common phenotypic feature, an erratic premature senescence (EPS) syndrome. Single tillers at any stage of development may start to senesce in response to environmental stress while other tillers stay green and healthy (Fig. 6G). Drought and mechanical injury of leaves are among factors that can trigger the syndrome. The expression intensity is dependent on the oat background, but the appearance itself is not. Because of the syndrome severity, the OMAd9.5 line could not be maintained. All F3 offspring died before seed set. The other lines are maintained as disomic addition lines, although there is inconsistent seed set in the plants. In most cases one or two tillers remained green and panicles developed normally. Other tillers died before heading. However, if the EPS syndrome is not expressed, the addition plant looks like normal oat with high fertility. Maize Chromosome 10 Additions Maize chromosome 10 was observed seven times among the F1 plants. It appeared only two times as single additions and five times in different combinations (Fig. 4). Six F1 plants died in early stages of development, but one F1 plant (OMAm10.1) could be kept alive. The plant is maintained under short-day conditions in several tiller clones. The plant has a grassy phenotype (Fig. 6H) and shows premature senescence, but not of the EPS syndrome type. When clone parts are shifted to long-day conditions, flowering is induced (Fig. 6I), but there has been no seed set to date. Maintaining the Oat-Maize Chromosome Addition Stocks All oat-maize chromosome addition lines maintained at the University of Minnesota are summarized in Figure 5. All stocks are grown in growth chambers under conditions described for regular oat in "Materials and Methods" with the exception of the lines OMAm1.1 (F1), OMAm1.7 (F1), and OMAd1.7 (currently F3). These chromosome 1 additions are being grown in short-day conditions only. All plants are isolated as single plants to guarantee self-breeding and to avoid cross breeding. F1 and F2 plants are isolated as single plants and their respective F2 and F3 offspring are harvested from single panicles. Beginning with the F3 generation, the plants are isolated as single plants and F 4 seeds are harvested and pooled from all panicles for every single plant. For OMAd1.7, line maintenance by harvesting seed from individual panicles will continue because of the sectoring activity observed. The chromosome 7 addition is propagated by individual panicles because of remaining instability for the added maize chromosome. The vegetative maintenance of tiller clones for OMAm1.1, OMAm5.9, OMAm5.17, and OMAm10.1 will continue until disomic addition offspring are generated for maize chromosomes 5 and 10. Experiments attempting to produce doubled haploid plants by application of colchicine are in progress. Genomic DNA of at least one line from each of the 10 different groups of oat-maize chromosome addition stocks has already been distributed to the scientific community and will continue to be available. To date, seeds are available for addition lines involving maize chromosomes 1, 2, 3, 4, 6, 7, 8, and 9. Besides the total of 35 single additions with all 10 maize chromosomes represented among them, two fertile double disomic addition lines are being maintained. One line is disomic for chromosomes 1 and 9 (OMAd1.12/d9.10) in Starter background. The second line is disomic for chromosomes 4 and 6 (OMAd4.6/d6.2), also in Starter background. DISCUSSION AND CONCLUSIONS Oat-maize addition lines are powerful tools for maize genomics
(Ananiev et al., 1997 A further utility of our addition set is the possibility to isolate
single maize chromosomes. By the application of flow cytometry, maize
chromosomes can be sorted and chromosome-specific DNA libraries become
feasible for all maize chromosomes. As the first example, maize
chromosome 9 has been isolated (Li et al., 2001 New insights of gene activation and silencing can be achieved by
analyses of maize gene activity in an oat host. The very special
situation that maize genes are transferred into oat enables special
inter-genomic effects to be observed and epistatic alleles to be
detected and defined. Muehlbauer et al. (2000) Addition lines allow the analysis of meiotic behavior such as
pairing, chromosome orientation, nondisjunction, and mobility as well
as chromosome structure in mitosis and interphase. GISH hybridization using labeled maize DNA enables
distinct visualization of an added maize chromosome pair in
an alien background. By using the oat-maize chromosome 9 addition line,
Bass et al. (2000) Maize genes and mobile elements may possibly be transferred into oat, which would facilitate gene tagging in oat, and eventually the combination of genetic information from plants with fundamental physiological differences, such as the long-day C3-type oat and the short-day C4-type maize, will raise new questions in the fields of plant physiology and biochemistry. This first complete set of oat-maize chromosome addition lines can significantly contribute to answering such questions.
Plant Material Single-plant selections of eight different oat (Avena
sativa) cultivars and experimental lines, including Starter,
GAF Park, Sun II, Kanota, Preakness, Stout, MN97201, and an
F1 (MN97201-1 × MN841801-1), were used in crosses
with maize (Zea mays) sweet corn hybrid Seneca 60, dent
corn inbred A188, or the experimental line bz1 mum-9,
which carries mobile transposable elements. Parental oat and maize
plants were grown with multiple planting dates in separate growth
chambers to obtain synchronized flowering time for crossing the two
species. Oat plants grow better at cool temperatures, i.e. 18°C to
20°C day, 14°C to 15°C night. Photoperiod is critical for the oat
lines used; all are of spring-habit type. Plants were grown under short
days (11-h light, 13-h dark) to favor vegetative growth and tillering
and after 6 to 8 weeks they were shifted to long days (16-h light, 8-h
dark) to induce flowering. Light of 300 to 400 µE m In Vitro Embryo Rescue Culture Primary florets of the upper two-thirds part of oat
panicles were hand-emasculated either by complete anther removal or by clipping florets. After 48 h these florets were pollinated with freshly shed maize pollen as described in Rines and Dahleen (1990) PCR To discriminate haploid oat plants without maize chromosomes
from those haploid oat plants with one or more retained maize chromosomes, 4- to 5-cm pieces from the first leaf blade of the F1 plants were cut and the genomic DNA extracted according
to Murray and Thompson (1980) To identify the added maize chromosome(s) in a maize-positive
F1 plant, the appropriate genomic DNA sample was
PCR-amplified in the presence of primers for maize chromosome-specific
SSR markers with the following program: 1× (94°C for 5 min), 35×
(94°C for 40 s SSR markers with known map coordinates were selected from the Maize Genome Database. Cytology To observe somatic chromosomes and interphase nuclei from
meristem cells, root tips (1.5-2 cm) were obtained by carefully removing the plant and soil from a pot containing a 3- to 4-leaf stage
maize-positive F1 plant and snipping off exposed root tips. Root tips were pretreated in a solution containing 0.05% (w/v) colchicine, 0.0025% (w/v) 8-hydroxyquinoline, and 1.5% (v/v)
dimethylsufoxide at room temperature for 3.5 to 4 h followed by
16-h ice water treatment to arrest cells at metaphase with good maize
and oat chromosome morphology. Root tips were fixed in a 3:1 (v/v)
mixture of ethanol and glacial acetic acid at room temperature for
2 d and stored at For chromosome counting, root tips were hydrolyzed in 1 N
HCl at 60°C for 12 min and stained in Schiff's reagent at room
temperature for 20 min (Feulgen and Rössenbeck, 1924 For GISH, root tips were washed by gentle shaking in distilled water
for 20 min and macerated in 45% (v/v) acetic acid at room temperature
for 5 min. The meristem cells were squashed in 45% (v/v) acetic acid.
GISH followed the general procedures of Pickering et al. (1997) A biometrical evaluation of differences in retention among the 10 maize chromosomes is not possible because of the limited number of maize chromosomes recovered in the F1 plants. Nomenclature and Conventions Because of the nature of the process of generating oat-maize chromosome addition plants, and the fact that each recovered line represents its own distinct retention event for every maize chromosome, we developed a nomenclature for identification of the addition lines as follows: Each line is named as OMAxy.z. OMA stands for oat, maize, and addition, respectively. The name includes three more alpha numericals: x, y, and z, with the z separated from the y by a period. The x would be an m in the case where the addition is monosomic. The x would be a d in the case where the addition is disomic. The x would be mt or dt in the cases where the additions are monotelosomic or ditelosomic, respectively. The y is the number of the maize chromosome that is added, thus 1 to 10 for the set. For a future possibility, the y may be a B, if a B chromosome of maize origin is added to the oat genome. The z is the identification number of the particular version of the maize chromosome present that traces back to the original single recovery event.
We thank Drs. Burle Gengenbach and John Gronwald (University of Minnesota) for reviewing the manuscript.
Received November 21, 2000; returned for revision December 14, 2000; accepted December 20, 2000. 1 This work was supported by the National Science Foundation (grant no. 9872650) and is a joint contribution of the Minnesota Agricultural Experiment Station and USDA-ARS. Mention of trademark or proprietary product does not constitute a guarantee or warranty by the University of Minnesota or USDA-ARS and does not imply approval over other products that also may be suitable.
2 Present address: Department of Crop and Soil Sciences, Oregon State University, 107 Crop Science Building, Corvallis, OR 97331-3002.
3 Present address: Trait and Technology Development, Pioneer Hi-Bred International, Inc., 7300 NW 62nd Avenue, P.O. Box 1004, Johnston, IA 50131-1004.
* Corresponding author; e-mail phill005{at}umn.edu; fax 612-625-1268.
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||