Plant Cell Reports

Plant Ceil Reports (1991) 10:152-155

(9 Springer-Verlag1991

Biochemical differences between carrot inbreds differing in plant regeneration potential R.P. Feirer and P.W. Simon Biological Department, St. Norbert College, De Pete, WI 54115 (RPF) and USDA-Agricultural Research Service, Department of Horticulture, University of Wisconsin-Madison, Madison, WI 53706 (PWS), USA Received February 13, 1991/Revised version received April 19, 1991 - Communicated by J. M. Widholm Abstract. Cultures derived from domestic carrot (Daucus carota L.) inbreds were found to vary with respect to regeneration potential as measured by the production of somatic embryos in suspension cultures. A number of biochemical parameters previously reported to distinguish embryogenic from non-embryogenic cultures of other species were measured in these carrot cell lines. Ethylene production was found to be inversely related to regeneration potential. The cell line producing the greatest number of somatic embryos exhibited the lowest rate of ethylene biosynthesis, even when grown on 2, 4-D-containing maintenance medium. A specific isozyme of acid phosphatase was associated with embryogenic calli. Proteins visualized by SDS-PAGE did not discriminate between embryo-forming and proliferating calli in all inbreds. Key words: Isozymes

Daucus - Regeneration - Ethylene - Protein -

Introduction The regeneration of plants from callus or suspension cultures is influenced by variables such as explant type, physical environment of the cultures, medium composition and the growth regulators employed. The importance of these factors was recognized relatively early in the development of the tissue culture systems, More recently, the genotype of the explant source has also been shown to have a major effect on the ease of plant regeneration. Genetic control of in vitro regeneration has been demon-strated in a wide variety of plants, including barley (Luhrs and Lorz 1987), maize (Hodges et al. 1986), tomato (Koornneef et al. 1987) and wheat (Carman et al. 1988). In a survey of seventy-six cultivars, genotypic variation of regeneration potential was also shown in alfalfa (Brown and Atanassov 1985). Earlier studies had indicated that two dominant genes control regeneration in this species (Reisch and Bingham 1980). Regeneration potential has also been reported to be under the control of two dominant genes in tomato, with regeneration having a high heritability (0.87) (Koornneef et al. 1987). In addition, it has recently been shown that not only regeneration potential, but also the ability of somatic embryos to germinate is under genetic control in soybean (Komatsuda and Ohyama 1988). Since the genetic influence on regeneration potential, specifically shoot-forming ability, had been shown in domestic Offprint requests to. R.P. Feirer

carrot (Trudeau-Spanjers 1987), somatic embryogenesis in several lines of carrot was the subject of a series of studies reported here. A goal of the work was to examine the regeneration potential, via somatic embryogenesis, of several contrasting carrot lines. The biochemical changes in cultures undergoing organogenesis or somatic embryogenesis have been the subject of considerable interest. A biochemical marker might be used to identify embryogenic cultures before morphogenesis was evident, to optimize culture conditions necessary for embryogenesis, to monitor the course of somatic embryogenesis, or perhaps to identify responsive explants. Several biochemical variables have been shown to discriminate between embryogenic and nonembryogenic tissues. Among these are proteins (Sung and Okimoto 1981), isozymes and ethylene. Everett et al. (1985) characterized maize calli types and demonstrated differences in secreted polysaccharides and isozymes, examination of esterase and G D H (glutamate dehydrogenase) isozymes revealed differences between embryogenic and shoot-forming cultures. Wann et al. (1987, 1989) employed biochemical markers to identify embryogenic and non-embryogenic conifer calli. Low levels of ethylene production, GSH and total reductants were found in the responsive callus type, but were significantly higher in non-embryogenic cultures. Described here are the results of the biochemical characterization of cultures derived from inbred domestic carrot lines. The variation of several biochemical parameters relative to the regeneration potential of the cultures or to active somatic embryogenesis was also tested. Materials and Methods Cell Culture. Carrot (Daucus carota L.) inbreds, B9304 and B493, from the USDA Carrot Genetics program at The Universityof Wisconsinwere used as explantsources. Calliwere initiatedfrom petiole explantsof aseptically germinated seedlings. Cultures were also obtained from root explantsof the F1 hybridof these two inbreds. Cultures were initiatedand maintainedon modified MS medium (Murashige and Skoog 1962) containing 2, 4-D and kinetin at 1.0 and 0.02 mg/1, respectively. The callus lines had been maintainedby monthlysubculture for at least 12 months at the beginning of these studies. Suspensioncultures derived from these lines were grown in 100 ml medium in 500 ml Erlenmeyer flasks incubated in darkness at 22 *C on a gyratoryshaker (100 rpm). Suspensionswere maintainedby transfer to fresh medium every 14 days. Embryo development was induced by inoculating screened (63/nn-250tan)and washedsuspensionsinto medium iackinggrowth regulators. Fifteen ml screw capped tubes containing2 ml medium were

153 inoculated at a densityof 0.5]A packed cell volume per ml of medium. The tubes were incubatedon a roller drum rotatingat approximately20 rpm. Embryo number was quantified by plating the contents of each tube in agar followed by visual scoring and counting on a Quebec colony counter. A structure was scored as an embryo if it exhibited a bipolar shape, had cotyledonary development, and was _>2mm in length (torpedo stage embryo or larger). During the routine maintenanceof calli from inbred B9304,a sector of one callus appeared to be composed of developing somatic embryos. This sector was isolated from the proliferating callus mass and maintainedas a separate line, designated B9304E. This line characteristicallyformed embryos and plantlets on medium containing 1 mg/1 2,4-D, which normally leads to the proliferation of callus and suppression of embryo formation.

Isozyme analysis. Callus tissues for isozyme analysiswere rinsed withdistilled water to remove adhering agar, weighed (fresh weight), and homogenized in 10:1 (vol:wt) of ice-cold 50mM HEPES buffer, pH 7.5, containing 1 mM PMSF (phenylmethyls-ulfonylfluoride, Sigma Chemical Company). After eentrifugationfor 5 min in an Eppendorf microfuge, protein was quantified in the supernatantby the Bradford dye bindingassay(Bradford 1976). Samples containing equal amounts of protein were loaded onto 0.5 mm thick 1% agarose gels containingAmpholines (LKB Corporation). Isoelectrie focusing in the pH range of 3.5-9.5 was carried out at 10~ for 30 min according to manufacturer'sinstruetions(Applicationnote 1818-A,LKB Corporation). The gels were then stained to visualize acid phosphatase (ltamaker and Snyder 1973), esterase (Everett et al. 1985), alcohol dehydrogenase (Everett et al. 1985) or peroxidase (Hamaker and Snyder 1973) activity. SDS-PAGEanalysis. Soluble proteins were extracted by homogenizing calli in ice-cold 50 mM ttEPES (pH =7.5) containing lmM PMSF as described above. After centrifugingfor 5 rain in an Eppendorf microfuge, protein in the supernatantwas quantified and loaded onto 7.5-15% gradient SDS-polyacrylamide gels (Laemmli 1970). Equal amountsof protein (2-3~) were loaded onto each lane and visualizedin the gel using a modified silver staintechnique (Oakley et ah 1980). Ethylene determination. Ethylene evolution from individual calli was determined by placing ca. 100 mg (fresh weight) tissue into 18 x 48 mm glass vialsor 3 ml Reacti-Vials (Pierce Chemical Company). Each vial,containing 5 ml or 2 ml medium, respectively, was loosely fitted with a polypropylene snap cap or screw cap fitted with card-board liners, respectively. Both of these caps allow free gas exchange. After 3 daysof incubation,the caps were replaced with rubber stoppers fitted with cylindricalrubber septa (#2-0668 LB-1, Supelco, Inc.). After an additional 24 hours of incubation, 1 ml of headspace gas was collected with a syringe through the septa. Ethylene was analyzed on a Packard (Model 428) gas chromatographequipped with a 6 ft Porapak Super Q packed columnand a flame ionization detector. The carrier gas (helium) flow rate was 20 ml/min, the oven temperature 45 ~ injector and detector temperatures 200 ~ Results W h e n maintained in the presence of 2,4-D, the physical appearance of the three carrot cultures employed in this study were substantially different. Macroscopically, the callus derived from B9304 was homogeneous and very friable, while that from B493 was nodular in both appearance and texture. Callus from the B9304E line was not only nodular, but also had the unique characteristic of containing embryos and small plantlets. Suspension cultures also exhibited these reproducible linespecific properties. Small, loosely organized d u m p s of cells characterized B9304 suspensions. In contrast, more compact d u m p s having smooth surfaces typical of proembryonic masses were found in B493 cultures. Suspension cultures of B9304E contained a variety of ceil and structure types. Both undifferentiated clumps and somatic embryos were present in these cultures throughout the subculture period with the somatic embryos being distributed among heart, torpedo and small plantlet stages. W h e n screened, washed and grown in the absence of growth regulators, embryo production by suspension cultures of the three lines reflected the appearance of the cultures on mainte-

nance medium. Line B9304E reproducibly produced a significantly greater n u m b e r of somatic embryos than B493 and its p a r e n t line, B9304 (Table 1). Cultures derived from B9304 always exhibited less regeneration potential than those from inbred B493, and occasionally did not form embryos when transferred to inductive conditions. W h e n grown in the presence of 2,4-D, ethylene production by calli was inversely correlated to the regeneration potential of the line. As illustrated in Table 1, ethylene production by the three lines was in the order B9304 > B493 > B9304E. Ethylene production by tissues grown under embryo-inducing conditions (growth regulators removed) was not significantly correlated to regeneration potential of the line. Table 1. Correlation of regeneration potential and levels of biochemical markers in different carrot lines grown in medium with (+GR) or without (-GR) growth regulators. Values within columns followed by the same superscript are not significantlydifferent as determined by Duncan's New Multiple Range Test (p < 0.05). Line

#Embryos (embryos/2 ml) -GR

B9304E B493 B9304

82" 25 b 2~

correlation: (#embryos vs x)

ethylene production (nl/g fw hr) +GR 1.6~ 6.4b 10.2a r = -0.99

-GR 1.6" 12.3 a 5.9 a -0.61

Examination of soluble proteins by S D S - P A G E was carried out using tissues grown on both proliferation (+ growth regulators) and induction (growth regulator free) culture media. As expected, the profile of proteins isolated from the three cell lines were similar, with most of the proteins being found in all of the tissues sampled (Figure 1). Several proteins were found to be present or more a b u n d a n t in specific lines, however. The bands identified as "a" and "b" appeared to be much more a b u n d a n t in extracts of B9304E than either B9304 or B493 when grown on the proliferation medium. The intensity of both of these bands increased in B9304 when grown in the absence of growth regulators. It appears, then, that the presence of these two bands is correlated with ongoing embryogenesis in cultures derived from B9304 inbred. Likewise, band c was found to be significantly more abundant in B493 than in B9304 and B9304E, although band c is somewhat more prominent in the latter tissues when they are grown in the absence of growth regulators. Overall, however, no protein was found to be consistently associated with all carrot cultures actively undergoing embryogenesis and no reliable protein marker of regeneration potential was evident. Although isozyme profiles of calli exhibited line (or genotype) dependent differences, no alcohol dehydrogenase, esterase or peroxidase isozyme bands could be related or correlated with regeneration potential or active embryogenesis (data not shown). Results obtained from gels stained to visualize acid phosphatase were more informative. O n proliferation medium, calli from B9304 and B493 differed with respect to a relatively dark staining anodic band (identified by arrow, Figure 2). Although absent in B9304 tissues, this band was present in B493 and in slightly greater amounts in extracts of B9304E. W h e n

154 grown on embryo-induction media (no growth regulators), this band was visualized in B9304 tissues and became more intense in tissues from the B9304E and F I lines. Since this isozyme band was found in all B9304-derived tissues actively undergoing embryogenesis and proliferating calli of line B493 containing proembryonic masses, it appears that it may be a marker of somatic embryo development. No significant correlation between regeneration potential and relative rates of protein synthesis, levels of GSH or total reductants have been observed (data not shown). Discussion

Fig. 1. SDS-PAGE examination of soluble proteins extracted from tissues grown on proliferation medium (+growth regulators) and regeneration medium (growth regulator free). The letters a, b and c next to the arrowheads identifybands discussed in the text, The black dots between several of the lanes indicate the position of these arrowheads.

Fig. 2. Isoeleetrie focusing gel stained to visualize acid phosphatase activity. Samples were obtained from tissues grown on proliferation medium (+growth regulators) and regeneration medium (growth regulator free). The arrow identifies an isozyme band associated with active embryo formation in calli derived from the inbred B9304. Equal a m o u n t s o f protein were loaded onto each lane (3.6/~g). Similarresults were obtained when samples representing equivalent tissue fresh weights were compared.

The cultures derived from inbred domestic carrot lines employed in this study clearly exhibited different regeneration potentials. The use of these cultures afforded the opportunity to study biochemical differences between lines related to their regeneration potential as well as changes associated with active somatic embryogenesis. As shown by SDS-PAGE, several soluble proteins were only found in cultures undergoing active somatic embryo development. Proteins identified as bands a and b were found in B9304E cultures, which produce embryos irrespective of the presence of growth regulators, and in B9304 tissues grown in medium lacking growth regulators. B9304 produced only a few somatic embryos when grown on this medium. Proteins found specifically in embryogenic callus, or somatic embryos, have been reported in carrot (Sung and Okimoto 1980, pea (Stirn and Jacobson 1987), and rice (Chen and Luthe 1987). Additionally, callus-specific proteins were also identified in several of these reports. The "embryo-specific" proteins identified in our work, however, were found in cultures derived from one inbred, B9304, and were not associated with somatic embryogenesis in cultures derived from a second inbred or the hybrid. It appears that the observation of "embryo-specific" proteins is genotype dependent, and these proteins are not markers of somatic embryogenesis across genotypes within a species. In the analysis of a limited number of isozymes, the presence of one acid phosphatase isozyme band in particular appeared to be associated with active carrot somatic embryogenesis. Lee and Dougall (1973) compared isozymes in proliferating and regenerating wild carrot cultures grown on inductive (2,4-D free) or prolifierative (2,4-D present) media and reported that GDH varied the most between the calli types. In the characterization of citrus cultures, embryogenic callus was found to have higher peroxidase activity than non-embryogenic callus, and a specific peroxidase isozyme was also associated with embryogenic callus (Kochba et al. 1977). The use of isozymes as markers of embryogenic cultures should be applied with caution since tissue or organ type, as well as the environment and culture medium, may influence isozyme expression in plant tissue. In our work, isozyme profiles have generally been found to be different in roots, seedling cotyledons and calli derived from the inbred plants (data not shown). In a study of ADH isozymes in carrot, cultured tissues were found to contain five rather than three isozymes found in intact tissues (Chourey and Widholm 1980). ADH activity was also apparently reduced by withdrawal of auxin from the culture medium. Since different isozymes are found in various plant organs, it is likely that isozymes reflect the stage of differentiation in the culture. Their regulation by the presence of growth regulators in the culture medium also complicates their use in the identification of embryogenic cultures or cultures having high regeneration potential.

155 In our experiments, the biochemical parameter most consistently related to regeneration potential and active somatic embryogenesis was ethylene biosynthesis. Low levels of ethylene production were clearly related to high regeneration potential in carrot cultures. Similar findings have been reported for tobacco and conifer calli, in which shoot-forming or embryogenic calli, respectively, produced less ethylene than non-regenerative calli (Huxter et al. 1981; Wann et al, 1987, 1989). Manipulation of ethylene levels has also been shown to dramatically affect tissue responses in vitro. Ethylene suppression of somatic embryogenesis and growth of carrot cultures has been demonstrated in cultures treated with Ethephon, an ethylene releasing compound (Tisserat and Murashige 1977). Treatment of embryogenic maize calli with norbornadiene and silver nitrate, both of which are ethylene antagonists, led to a 12-fold increase in plant regeneration from embryogenic calli (Songstad et al. 1988). Conversely, when the calli were treated with ACC (amino-cyclopropane-l-carboxylic acid), a precursor of ethylene synthesis, plant regeneration was significantly reduced. ACC treatments also led to the reduction of somatic embryo production from orchardgrass leaf explants (Songstad, et. al. 1989). Ethylene has also been shown to influence regeneration of sunflower (Robinson and Adams 1987). Treatment of explant sources with AVG (aminoethoxyvinylglycine), an inhibitor of ethylene biosynthesis, enhanced somatic embryo formation. The addition of ACC reversed these effects of AVG. Reports on the relationship between ethylene and in vitro morphogenesis are inconsistent, however. Although shootforming tobacco callus produced less ethylene than non-shoot forming callus, treatment of the cultures with ethylene-releasing agents or precursors enhanced shoot primordia formation (Huxter et al. 1981). Low levels of ethylene have also been utilized to induce root formation in cultured Digitalis, although higher levels did enhance the formation of undifferentiated callus and inhibited shoot formation (Perez-Bermudez, et al. 1985). It does appear that low rates of endogenous ethylene production are associated with tissues capable of or actively undergoing somatic embryogenesis. Also, the experimental reduction of ethylene levels generally leads to increased regeneration. A study involving inbred sun-flowers, suggesting that both the synthesis of and response to ethylene may be genotype dependent (Robinson et al. 1987), may relate to the line-dependent rates of ethylene production observed in our cultures. In conclusion, ethylene production was found to be the only consistent marker, across the genotypes examined, of regeneration potential and embryogenesis in carrot cultures. Highly regenerable cultures were characterized by low levels of ethylene production when grown on maintenance medium. Isozymes and proteins visualized by SDS-PAGE suffered limitations due to organ specificity and possible genotypic influences, respectively. Acknowledgments. The authors wish to thank J. Conkey and J. Carlson for valuable technical assistance. Portions of this work were used by RPF as partlalfulfillmentof the requirements for the Ph.D. degree at the University of Wisconsin-Madison. The University-IndustryResearch Programand The Instituteof Paper Chemistryprovided fellowship (RPF) and technical support, respectively.

References Bradford MM (1976) Anal. Biochem. 72:248-254 Brown DCW, AtanassovA (1985) Plant Cell Tiss. Org. Cult. 4:111-122 Carman JG, Jefferson NE, Campbell WF (1988) Plant Cell Tiss. Org. Cult. 12:83-95 Chen L, Luthe DS (1987) Plant Sci. 48:181-188 Chourey PS, Widholm J (1980) In Vitro 16:571-574 Everett NP, Wach MJ, Ashworth DJ (1985) Plant Science 41:133-140 Hamaker JM, Snyder EB (1973) U.S. Forest Service Research Note 50-151:1-8 Hodges TK, Kamo KK, Imbrie CW, Becwar MR (1986) Biotechnology4:219223 ttuxter T J, Thorp TA, Reid DM (1981) Physiol. Plant.53:319-326 Kochba J, Lavee S, Spiegel-Roy P (1977) Plant Cell Physiol. 18:463-467 KomatsudaT, Ohyama K (1988) Theor. Appl. Genet. 75:695-700 Koornneef M, Hanhar~ CJ, Martinelli L (1987) Theor. App]. Genet. 74:633641 Laemmli UK (1970) Nature 227:680-685 Lee DW, Dougall DK (1973) In Vitro 8:347-352 Luhrs R, Lorz tt (1987) Theor. Appl. Genet. 75:16-25 Murashige T, Skoog F (1962) Physiol.Plant. 15:473-497 Oakley BR, Kirsch DR, Morris NR (1980) Anal. Biochem. 105:361-363 Perez-Bermudez P, Cornejo MJ, Sequra J (1985) PlantCell Reports 4:188-190 Reisch B, Bingham ET (1980) Plant Set. Lett. 20:71-77 Robinson KEP, Adams DO (1987) Physiol. Plant 71:151-156 Robinson KEP, Adams DO, Lee RY (1987) Plant Cell Reports 6:405-409 SongstadDD, Duncan DR, WidholmJM (1988) PlantCell Reports 7:262-265 Songstad DD, Petraeek PD, Sams CE, Conger BV (1989) Plant Cell Reports 7:677-679 Stirn S, Jacobsen tt (1987) Plant Cell Reports 6:50-54 Sung ZR, Okimoto R (1981) Proc. Nat. Aead. Sci. 78:3683-3687 Tisserat B, Murashige T (1977) Plant Physiol.60:437-439 Trudeau-SpanjersM (1987) Masters Thesis. Universityof Wisconsin-Madison Wann SR, Johnson MA, Noland TL, Carlson JA (1987) Plant Cell Reports 6:39-42 Wann SR, Becwar MR, Nagmani R, Feirer RP, Johnson MA (1989) Trees 3:173-178