Dosage Effects on Gene Expression in a Maize Ploidy Series

Mei Guo, Doug Davis and James A. Birchler Division of Biological Sciences, University of Missouri, Columbia, Missouri 6521 1 Manuscript received August 24, 1995 Accepted for publication December 22, 1995

ABSTRACT Previousstudies on gene expression in aneuploidsrevealed numerous trancacting dosage effects. Segmental aneuploidy of each varied chromosomal region exhibited predominantly inverse effects on several target genes. Here, dosage regulation was examined in a maize (&a mays L.) ploidy series where the complete genomic complementis vaned. TotalRNA from leaf tissue of monoploid, diploid, triploid, and tetraploid plants (lx, 2X, 3X, and 4X, respectively) was analyzed for the expression of 18 genes. For most tested genes, the transcript level per cellis directly proportional to structural gene dosage; that is, on a per genome basis, thereis approximately equal expression among the four ploidies. Exceptional cases show a negative correlation of expression with ploidy aor positive correlation greater than expected from the structural gene dosage. These studies suggest that, in asgeneral, structural gene dosage increases in multiples of the monoploid complement, the absolute levelof gene expression per cell increases. In contrast, addition or subtractionof only a single chromosome arm tends to alter gene expression patterns extensively. The combined results of the euploid and aneuploid studies suggest that aneuploid effects result from an altered stoichiometry of the factors contributing to the mechanismsof gene expression.

HE gene dosage changes affecting structural gene In addition, when the dosage of a chromosomal seg- T expression have been studied in several species ment is varied, trans-acting effects, either positive (di- including human, mouse, Drosophila and plants rect) or negative (inverse), on genes elsewhere in the (BIRCHLER1979, 1981; BIRCHLERand NEWTON 1981; genome have been observed. We previously studied WHATLEYet al. 1984; EPSTEIN1986; REICHERT 1986; DEV- gene dosage effects of maizeaneuploids using B-A trans LIN et al. 1988; BIRCHLERet al. 1990; GUOand BIRCHLER locations to add or subtract single chromosome arms 1994). Changes in dosage include an aneuploid series relative to normal (GUO and BIRCHLER1994). Expres in which a chromosome or chromosomal segment is sion patterns of six genes were examined in a dosage added or subtracted relative to the genomic comple- series of 14 chromosomal segments. Transcript levels of ment or a ploidy series where a complete genomic in- each tested gene were affected by multiple chromo- crement is increased or decreased. When the dosage of somal regions. Each varied chromosomal region showed a structural gene is varied, typically two types of dosage trans-effects on several genes. Such modulations are pre- regulation on its expression are observed. One is a gene dominantly inverse (negative) effects on gene expres- dosage effect where there are increases or decreases in sion. When the dosage of a transacting inverse regulator expression in proportion to the gene copy number. of a target gene is varied simultaneously with the stmc- Gene dosage effects are often observed (e.g., GRELL tural gene, the dosage effect of the latter is cancelled 1962; EPSTEIN1986). Thisis particularly the case when by the former, resulting in the compensation described theaneuploid segments surroundingthe structural above (BIRCHLER1981; GUOand BIRCHLER1994). geneare small (BIRCHLER1981). Theother typeof Because of thepredominant lethality of ploidy dosage regulation is called dosage compensationin changes in animal systems, most ofthe studies on dosage which there is an equal expression regardless of the regulation ofploidy series have been conducted on copy number of the respective gene in aneuploids in- plants, although a few studies on yeast, Drosophila and volving significant cytological length. Studies of such animal cell lines have been reported (CIFERRIet al. 1969; dosage series from various organisms, especially Dro- PRIESTand PRIEST1969; LUCCHESIand RAWS 1973). sophila and plants, reveal that gene expression is often Among isozymesand proteins examined in various plant dosage compensated (BIRCHLER1979; BIRCHLERand species, some are positively or negativelyaffected by NEWTON 1981).This phenomenon is found with X ploidy changes, while most ofthem exhibit an increased chromosome linked genes,as well asgenes encoded on level as the ploidy level rises,that is, a gene dosage effect the autosomes of Drosophila (MULLER1932; DEVLINet (DEMAGGIOand LAMBRUKOS 1974;LEVIN et al. 1979; al. 1982; BIRCHLERet al. 1990). TIMKOet al. 1980; BIRCHLERand NEWTON1981). Mor- phologically documented effects withincreased ploidy in Curresponding author: James A. Birchler, Division of Biological Sci- ences, 117 Tucker Hall, University of Missouri, Columbia, MO 6521 1. plants include increased cell size and organ size while E-mail: [email protected] the vigor decreases beyond the tetraploid level (RHOADES

Genetics 142: 1349-1355 (April, 1996) 1350 M.Davis Guo, D. and J. A. Birchler

and MCCLINTOCK1935; BLAKESLEE1941; RHOADES and ploids and hexaploids fail to germinate orsurvive beyond the DEMPSEY1966). No previous studies have examined spe- early seedling stage. Plants of the ploidy series were transferred from the green- cific RNA levels in a ploidy series. house to the field after cytological confirmation by root tip In the present study, we investigated gene dosage chromosome counts. At the six-leaf stage the younger four effects further by varying the ploidy, where genomic leaves of each plant were harvested. The tissue was frozen in balance is maintained as compared with aneuploids. liquid nitrogen and stored at -80” until use. We sought to test whether the aneuploid effects were RNA isolation: Leaf tissue from single plants of each ploidy was used for RNA preparations. Total RNA was extracted us- cumulative, in principle, or whether a “balance” was ing amodified method of CONE et al. (1986). Tissue was operating as suggested by classical phenotypic observa- ground to a fine powder in liquid nitrogen with a mortar and tions. The transcript levels of 18 genes expressed in leaf pestle. Cells were lysed in 2.5-3.0 volumes of buffer con- tissue were analyzed inmonoploid, diploid, triploid and taining 0.1 M NaCl, 50 mM Tris-HC1 (pH 7.4),50 mM disodium tetraploid maize plants. Most genes exhibited a gene ethylenediaminetetraacetate (EDTA), 2% sodium dodecyl sulfate (SDS) and proteinase K (200 ,ug/ml). The supernatant dosage effect relativeto the ploidy level,although inter- was extracted twice with an equal volume of pheno1:chloro- esting exceptions occurred in which a strong positive form:isoamyl alcohol (100:99:1) and twicewith chloro- correlation or a negative correlation occurred between form:isoamyl alcohol (99: 1). Total RNA was precipitated in a gene expression and ploidy. final concentration of 2 M LiCl solution and reprecipitated with 2.5 volumes of absolute ethanol and one-tenth volume of 3.0 M sodium acetate (pH 5.0). MATERIALSAND METHODS Probe sources and Northern analysis: Except for the three well-characterized genes, Alcoholdehydrogenase 1 (Adhl),Alcohol Genetic stocks and production of a ploidy series: The dip- dehydrogenase 2 (AdhP) and Sucrosesynthase (Susl),those exam- loid (Oh 40B) and tetraploid (Alexander synthetic B) stocks ined were mapped isolates derived from a maize leaf cDNA were obtained from the Maize Genetics Cooperation. The library and were provided by the University of Missouri, Co- 1X -4X ploidy series was generated as follows: The monoploid lumbia (UMC) Maize RFLP laboratory. All clones for probe was produced by using the indeterminate gametophyte (ig)muta- synthesis were in or subcloned into theBluescript vector (Stra- tion, which causes the production of androgenetic haploids tagene) , which contains T3 and T7promoters. Antisense RNA probes were prepared by in vitro transcription of the respective when used as a female parent (KERMICLE 1969). The ig stock is homozygous for R-nj, which conditions anthocyanin devel- genes in such constructs using either the T3 or T7 promoter. opment in the anther, endosperm andembryo. This line was Five micrograms of total RNA from each sample was loaded crossed by the Oh 40B/elongate (el)diploid male (see below). in each lane and subjected to electrophoresis in 1.5% formal- Theandrogenetic haploid was selected based on the lack dehyde-agarose gels. RNAwas then capillary transferred to of pigmentation in the embryo. Triploids were produced by Biotrans nylon membrane (ICN) and cross linked by UV irra- crossing the tetraploid with the diploid. Each ploidy was con- diation. Blots were prehybridized for 6 hr and thenhybridized firmed by cytological examination of the chromosome num- with 3’‘P labeled antisense RNA of the respective genes. The ber in root tips. hybridization conditions were 60” in 5X SSC (1X SSC consists The homozygous elongate female produces normal haploid of 0.15 M NaCl, 0.015 M sodium citrate), 50% formamide, 1OX as well as diploid gametes (RHOADES and DEMPSEY1966). The Denhardt’s, 0.5% SDS, 10% dextran sulfate and 0.2 mg/ml el stock was crossed as a female parent with the Oh40B dip- salmon sperm DNA, for -16 hr. Theblots were washed at 75” loid. Normal (diploid embryos) as wellas defective kernels with 0.2X SSC and 0.05% SDS (BIRCHLERand HIEBERT1989). (triploid embryos) are produced. The normal kernels in the For each gene, triplicate loadings of each dosage series cross were grown and selfed. Females grown from theprogeny were made oneach gel. After probing with each specific gene, were crossed by the tetraploid stock of Alexander synthetic the blots were reprobed with antisense ribosomal RNA B. One-fourth of the females are expected to be el/el and (rRNA) as a measure of the loading of total RNA. To test that in these crosses ears would carry normal (tetraploids) and rRNA is a valid loading control, i.e., that constant rRNA level defective kernels (triploids). The plump kernels from such is produced per genome among theploidy series, total nucleic crosses are usually tetraploid, so they were selected and selfed acid isolated from leaf tissue was separated on an agarose gel for four generationsto recover a tetraploid elongate stock (Oh and stained with ethidium bromide. Negatives of film (type 40B/Alexander Synthetic B/el). This stock was crossed to the 55, Polaroid) were prepared, and the DNA and rRNA bands zg; R-nj stock in an attempt to produce androgeneticdiploids, were scanned with a laser densitometer. DNArRNA ratios but nonewere recovered. Therefore, the Oh40B/el stock was were then calculated according to the densitometric readings. used as the diploid. Consequently, the diploid and tetraploid No evidence of variation was found among theploidy series. stocks used in this experiment are Oh 40B/el and Oh 40B/ Data analysis: The radioactivity was quantitated using a Fuji Alexander Synthetic B/el, respectively. They are closely re- Phosphorimager. The specific RNA quantity of each lane was lated although they do not have an identical genetic back- normalized by calculating the ratio of mRNA/rRNA of the ground. Attempts were made to generate hexaploid and pen- phosphorimager readings. The means and standard errors taploid plants using the tetraploid elongate stock, which was of the respective normalized values from each ploidy were selfed and crossed by diploids, respectively; however, these calculated from three replicates. The mRNA levels of each higher ploidies are always associated with a defective endo- ploidy relative to that of the diploid were determined by tak- sperm, which impairs development. Seeds of the ploidy series ing theratios of the respective mean values of three replicates: were planted in vermiculite or Pro-mix soil in the greenhouse. monoploid/diploid, triploid/diploid and tetraploid/diploid, Approximately 300 such kernels for 5X selection and >ZOO respectively. Such ratios represent mRNAlevels relative to for 6X selection were planted. About 130 seedlings for 5X that of the diploid on a per genome basis because rRNA/ and 20 seedlings for 6X germinated and were screened by DNA ratios are constant among theploidy series. The values root tip chromosome counts, but none were above the 4X on a per cell basis were calculated by multiplying the per level. Therefore it appears thatin this background both penta- genome values by the respective genomic ratios of each ploidy M aize Ploidy Series Ploidy Maize 1351 to diploid: Le., 0.5 for monoploid, 1.0 for diploid, 1.5 for served with the ploidy series is in contrast to what was triploid and 2.0 for tetraploid. Using csulb as an example, found in the aneuploiddosage series, where the expres- the per genome values from monoploid to tetraploid0.97, are 1.OO, 1.01 and 1.OO, respectively; their per cell values are0.49 sion of most genes located on the varied chromosome (0.97 X 0.5), 1.00 (1.00 X 1.00), 1.52 (1.01 X 1.5) and 2.00 arm was dosage compensated. Moreover, the multitude (1.00 X 2.00), respectively. of trans-acting aneuploid effects does not appear to be cumulative. In contrast to the standard dosage effects, there were RESULTS several interesting exceptions among the 18 genes ana- Numerous dosage effects were found in the previous lyzed. The first one is csu 5, a gene with homology to study of aneuploid dosage series of 14 chromosomal thiol protease, on chromosome 7. The transcript level segments (Guo and BIRCHLER1994). This work led us of csu 5 decreases as the ploidy increases (Table 1 and to investigate dosage regulation of gene expression in Figure 1). If the reduction is inversely proportional to a euploid series, consisting of monoploid, diploid, trip- the number of genomes, on a per cell level, taking the loid and tetraploid maize plants to determine whether diploid as loo%, the predicted value of the monoploid such effects are additive or nonadditive in general when will be a twofold increase (200%, inverse of 1/2), the collective regions of thegenome are varied to- whereas in the triploid andtetraploid, it wouldap- gether. Northern analyses were used to determine the proach 67% (inverse of 3/2) and 50% (inverse of 4/ transcript levels of 18 different genes including 15ran- 2), respectively.However, in the higher ploidies, the dom cDNA clones from maizeleaf tissue, as well as actual decrease is more extreme with the relative levels Adhl, Adh2 and Susl. To confirm that ribosomal RNA at 48 and 28% of the diploid, respectively. In other could be used to standardize equal loadingof total RNA words, the reduction is more than that expected from on Northern blots, total nucleic acid from the ploidy an inverse effect due to the increasing ploidy level. series was quantified and compared. The ratios of Another exceptional pattern was exhibited by csu ?l, rRNA/DNA were calculated and the following values which is on chromosome 8. There is a striking increase were obtained for each ploidy: 0.97 (monoploid), 1.04 in the amount of mRNA from 1X to 4X. The increased (diploid), 0.95 (triploid) and 1.08 (tetraploid). These messenger RNA level pergenome with increasing results indicate that approximately equal rRNA levels ploidy was 68% (lx), 100% (2X), 251% (3X)and are presentper genome among theploidy series. There- 214% (4X), or per cellvalues of 34% (lx),100% (2X), fore, rRNA could be used as a valid internal loading 377% (3X) and 428% (4X).These levels exceed those control. predicted from a direct gene dosage effect (Figure 1 Values from the Northern analysis (see Table 1) are and Table 1). A direct gene dosage effect will give an expressed as relativetranscript levels ofeach gene from equal transcript level per genome among ploidies, or each ploidy to that of the diploid and are normalized per cell valuesof 50% for monoploid,100% for diploid, to a per genome basis as well as a per cell basis. If a 150% for triploid and 200% for tetraploid. gene is expressed in proportion to its dosage, an equal A thirdexceptional expression pattern was exhibited expression value will be measured per genomeregard- by Susl, which is encoded on chromosome 9. The mes- less of the ploidy level. To obtain the per cell level, the senger RNA level of Susl exhibits a curious odd/even expression in each ploidy was multiplied by the number effect as illustrated in Table 1 and Figure 1. That is, in of genomes present in a cell, that is, 1X, 2X, 3X and themonoploid and triploid the transcript levelsin- 4X in monoploid, diploid, triploid and tetraploid, re- crease dramatically relative to the diploid butthe spectively (because the value of the diploid is standard- mRNA levels of the diploid and tetraploid are not sig- ized as 1.00, the per cell values of each ploidy were nificantly different from each other. The magnitude of calculated in proportion to the diploid; see MATERIALS the increase is so extreme that there is a nearly three AND METHODS). Results from the study indicate that the fold higher mRNA level per genome in the monoploid transcript level per genome of most genes is approxi- and 6.7-fold higher in the triploid as compared with mately equal among differentploidies or, on a per cell the diploid and tetraploid. Although less dramatic, two basis, the transcript level of these genes increases pro- other genes, csu 28 (unmapped) and csu 116 (chromo- portionally with increasing ploidy. Thus, most genes some 6),also display similar patterns of the odd/even display gene dosage effects, although variations are ob- effect (Table 1). The transcript level of csu 28 has a served withsome of the clones. A Northernblot probed 1.36-fold increase in the monoploid and a 2.44fold in- with csu 16, a clone encoding a proteinwith homology crease in the triploid as compared with that of the dip- to NADP malic enzyme, is presented in Figure 1 as an loid; csu 116 expresses 1.41-fold higher mRNA level in example of the gene dosage effect of ploidy. Because the monoploid and twofold higher in the triploid as approximately equal amountsof total RNA were loaded compared with the diploid. and rRNA/DNA ratios are constant amongploidies, the DISCUSSION mRNA levelson Northernblots represent a per genome For gene dosage studies it is desirable to have a uni- measure. Thepredominant gene dosage effect ob- form genetic background through the dosage series. As 1352 M. Guo, D. Davis and J. A. Birchler

TABLE 1 The mRNA levels of 18 genes expressed in leaf tissue of four ploidies relative to the diploid

~~ csu5 (7") csuI6 (3) csul7 (2) Thiol protease NADP malic enzym 31k ribonucleo$rotein Ploidyb Mean" SEd Ploidy Mean SE Ploidy Mean SE

~~ 1 2.00 0.05 (1.00) 1 0.97 0.17 (0.49) 1 0.84 0.13 (0.42) 2 1.00 0.04 (1.00) 2 1.oo 0.16 (1.00) 2 1.oo 0.12 (1.00) 3 0.32 0.00 (0.48) 3 1.01 0.12 (1.52) 3 1.29 0.14 (1.94) 4 0.14 0.02 (0.28) 4 1.oo 0.10 (2.00) 4 1.03 0.09 (2.06) csu26 (5) mu28 (-) csu30 (3) ATP/ADP translocator Ribosomalprotein S22 ATPaseVacuolar proteolipid subunit

1 1.10 0.02 (0.55) 1 1.36 0.02 (0.68) 1 1.44 0.03 (0.72) 2 1.00 0.03 (1.00) 2 1.oo 0.01 (1.00) 2 1.oo 0.01 (1.00) 3 1.56 0.03 (2.34) 3 2.44 0.06 (3.66) 3 1.10 0.02 (1.65) 4 1.25 0.01 (2.50) 4 1.81 0.01 (3.62) 4 1.18 0.04 (2.36) csu31 (8) (4) ~~91-I csu77 (I) Unknown Malate dehydrogenase Unknown

1 0.68 0.01 (0.34) 1 1.47 0.02 (0.74) 1 1.19 0.02 (0.60) 2 1.oo 0.03 (1.00) 2 1.00 0.03 (1.OO) 2 1.oo 0.01 (1.00) 3 2.51 0.08 (3.77) 3 1.24 0.06 (1.86) 3 1.23 0.02 (1.85) 4 2.14 0.06 (4.28) 4 1.52 0.04 (3.04) 4 1.02 0.01 (2.04) c~u91-2(4) csu9? (9) csu96 (3) Unknown Unknown Thiol protease inhibitor

1 0.99 0.05 (0.50) 1 1.65 0.35 (0.83) 1 1.23 0.05 (0.62) 2 1.oo 0.01 (1.00) 2 1.oo 0.22 (1.00) 2 1.oo 0.03 (1.00) 3 1.34 0.01 (2.01) 3 1.26 0.29 (1.89) 3 0.96 0.03 (1.44) 4 1.11 0.03 (2.22) 4 0.81 0.20 (1.62) 4 1.13 0.03 (2.26)

~ ~ csull6 (6) csu137 (5) csul40 (IO) Elongation factorElongation la MADS box gene Glyceraldehyde 3P dehydrogenase

1 1.41 0.04 (0.71) 1 0.92 0.04 (0.46) 1 0.99 0.04 (0.50) 2 1.oo 0.01 (1.00) 2 1.oo 0.06 (1.OO) 2 1.oo 0.01 (1.00) 3 2.01 0.03 (3.02) 3 1.45 0.07 (2.18) 3 1.43 0.01 (2.15) 4 1.46 0.02 (2.92) 4 0.69 0.02 (1.38) 4 1.15 0.02 (2.30) Adh l (I) Adh2 (4) Susl (4) Adh2 (I) Adhl (9) Alcohol dehydrogenase 1 Alcohol dehydrogenase 2 Sucrose synthase 1

1 1.40 0.16 (0.70) 1 0.93 0.14 (0.47) 1 2.94 0.04 (1.47) 2 1.oo 0.10 (1.00) 2 1.oo 0.15 (1.00) 2 1.oo 0.03 (1.00) 3 0.75 0.06 (1.13) 3 0.93 0.13 (1.40) 3 6.71 0.32 (10.07) 4 0.98 0.06 (1.96) 4 1.15 0.15 (2.30) 4 0.87 0.03 (1.74) Mean values and SE are presented on a per genome basis with the corresponding per cell levels in parentheses. a Indicates the chromosome where each gene is mapped. bPloidy 1, 2, 3, and 4 refers to monoploid, diploid, triploid, and tetraploid, respectively. Mean is an average value of the relative mRNA levels from three replicates. Mean values that are in bold face are significantly (P< 0.01 in t tests) different from that of diploids. 'SE, Standard error of three replicates. we were unsuccessful in generating diploid plants with is partially related to the tetraploid. If this potential heter- the Alexander synthetic B stock crossed onto zg females, ogeneity affected expressionpatterns, it would be found a modified approach was taken (see MATERIALS AND as pronounced differences between the monoploid/dip METHODS). The monoploid and diploid are derived from loid lines and the triploid/tetraploid lines. No such pat- the Oh 40B/eZ stock. The tetraploid is derived from a tern is obvious in the data. Therefore, we do not believe selfed progeny from crosses of Oh 40B/el X Alexander that the minor differences in background have a pro- synthetic B tetraploid. Therefore, the monoploid and found effect on the major results, but one should con- diploid are derived from the same genetic stock, which sider this situation in interpretation of differences. Maize Ploidy Series 1353

csu 16 csu 5 csu 31 sus 1

NADP malic enzyme Thiol protease unknown Sucrose synthase Ploidy 12341234 1234 1234 ”WW

”

rRNA +

FIGURI:.1.-The mRNA levels of four genes representing different response patterns in leaf tissue of four ploidies. Blots were first probed with the respective genes and then with rRNA as a loading control.

It should also be noted that there is a certain degree genes. In aneuploids, chromosomal imbalance would of aneuploidy in triploid and tetraploid maize (RAN- change the stoichiometry of these factors, resulting in DOLPH 1935). That is, a tetraploid plant often consists altered geneexpression patterns. In the euploid series, of moreor less than exactly40 chromosomes. The the dosage of the genomic complement is changed plants used in this study were cytologically confirmed proportionally and the stoichiometric relationship are as exact multiples of the monoploid number.However, presumably maintained. Therefore, the level of gene a tetraploid with 40 chromosomes may not have four expression will be increased or decreased according to setS of each chromosome; instead, it may have, for ex- the ploidy variation. ample, four sets of chromosomes 2-8 plus three chro- Nevertheless, in the present study of the 18 genes mosomes 10 and five chromosomes 9 (8 IV+III+V). analyzed, several do not followthis trend and show Such aneuploidy of triploid and tetraploid plants might remarkable expression patterns. The transcript level of also contribute to minor deviations from the expected the thiol protease-like gene, csu 5, decreases dramati- gene dosage effect. cally with increasing genomic equivalents. This result Results from this study demonstrate that increasing suggests that the absolute level of a regulatory product gene dosage from one to four via a ploidy series in- responsible for this effect negatively influences the level creases the expression ofmost genes proportionally. of csu 5 mRNA. In Brassica ohacea, NAKAI(1977) also Polyploids innumerous plant species have shownhigher noted decreased activity for some esterase isozymes in levels of enzyme activity or proteins than diploids (DE- the tetraploid in comparison with its diploid. There MAGCIO and LAMBRUKOS 1974; NAKAI1977; LEVINet al. were even more extreme cases in which some isozyme 1979; TIMKOet nl. 1980). Similar observations were also bands found in diploids were absent in the tetraploids. made withyeast (CIFERRIet al. 1969) and Drosophila A consistent decrease in esterase isozyme activity with (LUCCI-IESIand RAMIS 1973). In the fibroblast-like cell increased ploidy was also observed in monoploid, dip lines of rat, the rate of collagen synthesis was correlated loid and tetraploid individuals of Ricinus (TIMKOet al. with the ploidy (PRIEST andPRIEST 1969). Although ex- 1980) and peroxidase isozyme activityin a ploidy series ceptions exist, such direct correlations of gene expres- of the gametophyte and sporophyte of ferns (DEMAG sion with the copy number of the genome in a ploidy CIO and LAMBRUKOS 1974). It was suggested that an series are the predominant situation. increased dosage of a regulatory gene(s) may have In contrast, results from our earlier studies on the played a role in determining theactivity of the affected transcript levels of six genes examinedin an aneuploid isozyme bands (NAKAI 1977). On the other hand, for dosage series of 14 chromosomal segments of maize certain genes like csu 31, the positive correlation of revealed a high level of modulation(GUO and transcript level to ploidy can go beyond the level pre- BIRCHLER1994). For instance, in the case of Adhl and dicted from a strict gene dosage effect. MITRA and Adh2, within the ploidy dosage series, no significant BHATIA(1971) also observed such “super” positive cor- variation of the transcript amount per genome was relation of several isozyme activities(ADH, malate dehy- found. However, in the aneuploid dosage series, the drogenase and glutamate dehydrogenase) in a ploidy expression of both geneswas dosage compensated and series including diploid (AA), tetraploid (AABB), and modulated by several chromosomal regions in trans. hexaploid (AABBDD)wheat and an octaploid of Tritici- Our hypothesis is that the effect of dosage regulation neae (AABBDDRR) derived from wheat and rye. These on gene expression is the result of the stoichiometric observations further indicate an interaction of struc- interactions of multiple dosage sensitive trans regula- tural gene dosage and theirregulatory genes in produc- tory factors among themselves and with their target ing the net effects on gene expression. 1354 M. Guo, D. Davis and J. A. Birchler

The interesting expression pattern exhibited by Susl ever, further increases in autopolyploidy result in plants is that the transcript level is elevated nearly threefold with decreased vigor. Based upon our results, these ef- in the monoploid on a per genome level and more fects could be brought aboutby the exceptional cases in than sixfold in the triploid compared with the diploid which gene expression decreases with increasing ploidy, and tetraploid. A similar expression pattern is also rep- although the relative changes of gene products or the resented by two other genes, csu 28 and csu 116. This effects of different cell sizeon developmental processes typeof result is reminiscent of the odd/even effect might also contribute. found with B chromosomes in rye and maize (JONES We thank L. C. HANNAH,M. M. SACHS,R. PHIILIPSand the UMC and REES 1982). In these two cases, the levels of total maize RFLP lab for providing clones. Research was supported by a nuclear protein and RNA (KIRKand JONES 1970; AYO- grant from the Departmentof Energy Biosciences Division. NOADU and REEs 1971) and also the straw weight and the number of tillers per plant (MUNTZING1963) were elevated when an even number of B chromosomes were LITERATURE CITED present in the cell and reduced to a lower level when AYONOAOU,U., and H. REES, 1971 The effects of B chromosomes there was an odd number of B chromosomes present. on the nuclear phenotype in root meristems of maize. Heredity The direction of the effectsbetween the B chromo- 27: 365-383. BIRCHLER,J. A,, 1979 A study of enzyme activities in a dosage series somes and the ploidy series is opposite, but thetwo have of the long armof chromosome onein maize. Genetics 92: 121 1 - in common an unusual response due to chromosome 1229. number. Itis not known whatfactors are involved in this BIRCHLER,J. A., 1981 The genetic basis of dosage compensation of Alcohol dehydrogenase-l in maize. Genetics 97: 625-637. phenomenon or whether one or more A chromosomes BIRCHLER,J. A,, and J. C. HIEBERT,1989 Interaction of the Enhancer might produce such an effect as suggested by the Susl of white-apicot with transposable element alleles at thewhite locus result. It is of interest to note that Susl is the only one in Drosophila melanogaster, Genetics 122 129-138. BIRCHLER,J. A,, and IC J. NEWTON,1981 Modulation of protein of sixgenes examinedin our previous aneuploidy study levels in chromosomal dosage series of maize: the biochemical that exhibited effects that fell outside the limits of an basis of aneuploid syndromes. Genetics 99: 247-266. inverse or direct correlation with dosage (GUO and BIRCHLER,J. A., J. C. HIEBERTand IC PAIGEN,1990 Analysis of autosomal dosage compensation involving the Alcohol dehydrogenase BIRCHLER1994). For example, a transacting direct ef- locus in Drosophila melanogaster. Genetics 124 677-686. fect will reduce the mRNA level of a two dose endo- BIAKESLEE,A. F., 1941 Effect of induced polyploidy in plants. Am. sperm to 67% (2/3) and enhance that of a four dose Ndt. 75: 117-135. CIFERRI,O., S. SORAand 0.TIBONI, 1969 Effects of gene dosage endosperm to 133% (4/3) of the normal three dose on tryptophan synthase activity in Saccharomyces cermisiae. Genet- endosperm. However, the transcript level of Susl (on ics 61: 567-576. chromosome 9) was 58 and 246% in endosperm tissue CONE,K. C., F. A. BURR and B. BURR,1986 Molecular analysisof the maize anthocyanin regulatory locus Cl. Proc. Natl. Acad. Sci. containing two and fourdoses of the shortarm of chro- USA 83: 9631-9635. mosome 4, respectively. An interesting future study DEMAWIO,A. E., and J. LAMBRUKOS,1974 Polyploidy andgene would be to examine the exceptional ploidy cases in dosage effects on peroxidase activity in ferns. Biochem. Genet. 12: 429-440. the aneuploidy series to determine whether theyall DEVLIN,R. H., D. G. HOLMand T. A. GRIGLIATTI,1982 Autosomal exhibit unique aneuploid effects. If present, the odd/ dosage compensation in Drosophila melanogaster strains trisomic even effect could then be experimentally approached for the left arm of chromosome 2. Proc. Natl. Acad. Sci. USA 79: 1200-1204. using chromosomal manipulation techniques available DEVLIN,R. H., D. G. HOLMand T. A. GRIGLIATTI,1988 The influ- in diploid maize. ence of whole-arm trisomy on geneexpression in Drosophila. Ge- In conclusion, gene regulation in response to seg- netics 118: 87-101. EPSTEIN,C. J., 1986 The Consequences of Chromosome Imbalance. Cam- mental dosage variation results from a combined effect bridge University Press, New York.

of the structural genes and the interactive relationship GREIL,E. H., 1962 The dosage effect of ma-l+ and TY+ on xanthine among dosage sensitive modifiers. When this relation- dehydrogenase activity in Drosqbhila melanogaster. Z. Verebungsl. 93: 371-377. ship is maintained in euploidy, gene dosage effects are Guo, M., and J. B. BIRCHLER,1994 Trans-acting dosage effects on often observed. When this system is altered, as in the the expression of model gene systems in maize aneuploids. Sci- aneuploid condition, modifications of expression pat- ence 266 1999-2002. JONES,R. N., and H. REES, 1982 B Chromosomes. Academic Press, terns result. In most multicellular eukaryotes, aneu- New York. ploidy has a more severe effect upon phenotype than KERMICLE, J. L., 1969 Androgenesis conditioned by a mutation in altering ploidy level. Our molecular results are in paral- maize. Science 166: 1422-1424. KIRK, D., and R. N. JONES, 1970 Nuclear genetic activity in Bchro- lelwith these observations. We havepreviously pro- mosome rye, in terms of quantitative interrelationships between posed that the reductions in gene expression found in nuclear protein, nuclear RNA and histone.Chromosoma 31: 241-254. both monosornics and trisomics are rate limiting on LEVIN,D. A,, A. M. TORRESand M. LEW, 1979 Alcohol dehydroge- growth and could be the molecular basis of aneuploid nase activity indiploid and autotetraploidPhlox. Biochem. Genet. syndromes (BIRCHLERand NEWTON 1981; GUO and 17: 35-43. BIRCHLER1994). In a ploidy series, gene expression LUCCHESI,J. C., and J. M. RAWLS, 1973 Regulation of gene function: a comparison of enzyme activity levelsin relation to gene dosage patterns are more constant as are the phenotypes, al- in diploids and triploids of Drosophilamelanogaster. Biochem. though each ploidy has characteristic properties. How- Genet. 9: 41-51. Maize Ploidy Series 1355

MITRA,R., and C. R. BHATIA,1971 Isozymes and polyploidy. Genet. from mouse fetuses with trisomy 19 after DEAE sepharose chrc- Res. 18: 57-69. matography. Genet. Res. 47: 193-197 MULLER,H. J., 1932 Further studies on the nature and causes of RHOADES, M. M., and E. DEMPSEY,1966 Induction of chromosome gene mutations. Proc. Int. Cong. Genet. 1: 213-255. doubling at meiosis by the elongate gene in maize. Genetics 54: MUNTZING,A,, 1963 Effects of accessory chromosomes in diploid 505-522. and tetraploid rye. Hereditas 49: 371-426. RHOADES,M. M., and B. MCCLINTOCK,1935 The cytogenetics of NAKAI, Y., 1977 Variations of esterase isozymes and somesoluble maize. Bot. Rev. 1: 292-325. proteins in diploids and their induced autotetraploidsin plants. TIMKO,M. P., A. C. VASCONCELOSand D. E. FAIRBROTHERS,1980 Jpn. J. Genet. 52 171-181. Euploid in Ricinus. I. Euploidy and gene dosage effects on cellu- PRIEST,R. E., and J. H. PRIEST,1969 Diploid and tetraploid clonal lar proteins. Biochem. Genet. 18: 171-193. cells in culture: gene ploidy and synthesis of collagen. Biochem. WHATLEY,S. A,, C. HALL,A. N. DAWSONand L. LIN,1984 Alterations Genet. 3: 371-382. in the relative amounts of specific mRNA species in the developing RANDOLPH,L. F., 1935 Cytogenetics of tetraploid maize. J. Agric. human brain in Down’s syndrome. Biochem. J. 220 179-187. Res. 50: 591-605. REXCHERT,G. H., 1986 Two-dimensional gel analysis of proteins Communicating editor: W. F. SHERIDAN