Introduction

Plant tissue culture systems allow for the rearing of whole plants, organs, tissues, or cells under controlled aseptic condi- tions in the laboratory. The tissue culture system supplies all nutrients, energy, and water necessary for plant or explant growth through the basal medium. In addition, controlled in- cubation conditions provide optimized light and temperature settings to promote growth. Plant development can then be manipulated by the addition of plant growth regulators (either

natural phytohormones or synthetic versions) at particular stages of growth or maturation.

This current review is structured to follow the general se- quence of plant tissue culture protocols, beginning with plan- ning and preparation, and commencing through initiation, proliferation, and regeneration steps. Thus, first, a brief review of the considerations for choice of the basal medium and fun- damental incubation conditions is presented. Second, com- monly used plant growth regulators (synthetic versions of phytohormones) and how they may achieve their physiologi- cal effects are briefly surveyed. Third, the classical phytohor- mone and developmental models for manipulating vegetative

development in vitro are reviewed. In the final section, more

detailed developmental models are explored, which demon-

* Gregory C. Phillips gphillips@astate.edu

¹ College of Agriculture, University of Arkansas Division of Agriculture Experiment Station, Arkansas State University, State University, P.O. Box 1080, Jonesboro, AR 72467, USA

² Molecular BioSciences Program, Arkansas State University, State University, P.O. Box 837, Jonesboro, AR 72467, USA

³ College of Sciences and Mathematics, Arkansas State University, State University, P.O. Box 1030, Jonesboro, AR 72467, USA

strate additional concepts and principles of basal medium and plant growth regulator manipulation, as related to in vitro plant regeneration.

In order to provide relevant context for the discussion of choice of basal medium formulation, the citations related to plant tissue culture regeneration media compiled by Herman (2015) over the period of 2011 to 2015 were surveyed. Although some of these citations were sourced from scientific meeting abstracts, the majority of these citations were  

gathered from the primary original literature. Focusing on the primary original literature citations, there were 123 useful sources, which may or not be comprehensive. Nevertheless, these citations represent a recent 5-y snapshot of plant tissue culture medium usage as documented in the literature and serve as a basis for discussion of recent and current practices in the discipline.

Fundamental Considerations: Basal Medium and Incubation Conditions

Two distinct trends were noted among this recent set of cita- tions (Herman 2015). First, efforts to optimize or customize the basal medium for specific species or genotypes, and for specific developmental steps, such as shoot multiplication or rooting, were ongoing. Response surface methodology, design of experiments (DOE) methodology, and neuro-fuzzy logic software programs were introduced during this time period for basal medium customization. Most of these studies used MS basal medium (Murashige and Skoog 1962) at the start of optimization, but as discussed in Greenway et al. (2012), sometimes MS medium components were so inadequate that optimization was not achievable. In those cases, the use of BABI, a B5 (Gamborg et al. 1968) formulation modified at Arkansas Biosciences Institute (Greenway et al. 2012), might be recommended as an alternative starting point for optimiza- tion. The BABI basal medium utilizes a somewhat different set of macronutrient salts than does MS, separating the nitrate from ammonium in different salts, so customization for a par- ticular species or genotype can be fine-tuned to lead to differ- ent endpoints. Second, tissue culture protocols were increas- ingly utilized for high-throughput phenotyping, for example, screening large numbers of in vitro regenerated plants or in vitro germinated seedlings for a particular transgenic or mutant phenotype, using an automated multi-spectral imaging system such as a ScanAlyzer (LemnaTec GmbH, Aachen, Germany). Published, preferably standardized, protocols for the particular species or genotypes under study should be used for high-throughput phenotyping whenever possible, to allow better comparison of datasets across research groups. Should published or standardized protocols not be available for a par- ticular application, one may be developed using the practices presented in this present review.

Although prepackaged MS, B5, and modifications of these basal media are available commercially, many plant tissue culture labs prepare media using groups of the individual chemical components as stock solutions. Macronutrient salts (N, P, K, Mg, and S) can be prepared together in a 10× or 100× stock solution (Gamborg et al. 1976; Gamborg and Phillips 1995). Calcium salts are segregated into a separate stock so- lution, because insoluble calcium phosphate salts are formed when Ca is included in the macronutrient stock. Iron

ethylenediaminetetraacetic acid (FeEDTA) is made as a sepa- rate stock, using equimolar amounts of FeSO₄·7H₂O and Na₂EDTA, and FeEDTA stock must be autoclaved to force the maximum formation of FeEDTA before use, because EDTA can react with other elements, especially certain micronutrients. Micronutrient salts (B, Mn, Zn, Mo, Cu, Co, and I) can be grouped into a single 100× or 1000× stock solution. Macro- and micronutrient salt stocks can be autoclaved for storage or frozen in aliquots. Double-distilled or reverse osmosis water is preferred for preparation of chem- ically defined media.

Vitamins and some organic compounds (typically thiamine, pyridoxine, nicotinic acid, and myo-inositol) can be grouped together into a single stock solution (100×), but this stock solution should be refrigerated for storage and should not be autoclaved prior to medium preparation. Growth regulators can also be made into individual stock so- lutions, at 1 mg mL⁻¹, and stored in the refrigerator (these organic compounds should not be autoclaved, except in the final prepared medium). Some heat-sensitive organic com- pounds may be added to the autoclaved medium after filter sterilization. Carbohydrates, such as sugars, typically are added dry to the basal medium, but some labs prefer to filter-sterilize the carbohydrate because autoclaving can lead to breakdown or carmelization. Gelling agents, such as differ- ent types of agars, agarose, gellan gum, or calcium alginate, do have distinct effects on plant tissue culture responses. When it comes to medium preparation, the most important principle is consistency; if a customized tissue culture protocol identifies a particular gelling agent, the specific type and brand should be used for all experiments of a given kind.

Commonly used basal media Not surprisingly, the most widely used plant tissue culture basal medium is MS (Murashige and Skoog 1962; see Table 1) and modifications thereof, such as half-strength MS (½-MS); MS-based media were used in 82% of the 5-y set of citations (Herman 2015). While at one time in the past, B5 medium (Gamborg et al. (1968; Table 1) or its modifications were widely used, in this 5-y set of citations, B5-related media were used in only 5% of the cases. Numerous citations identify the use of MS salts with B5 vitamins. Schenk and Hildebrandt medium (SH; 1972) was mentioned only once (< 1%); it and other specialty basal media are documented in Gamborg and Phillips (1995). Woody plant basal media (see Table 1) such as Woody Plant Medium (WPM; Lloyd and McCown 1980) and Driver and Kuniyuki Woody plant medium (DKW; Driver and Kuniyuki 1984) were used in 6% of the 5-y set of citations. Another 7% of the citations used ½-MS or MMS (modified MS; ½-macro MS salts + full-strength micro MS salts + B5 vitamins) for woody plant applications, so altogether woody plant basal media comprised a total of 13% of the citations. Half- strength MS was used in another 6% of the citations for

Table 1. The nutrient concentrations of commonly used plant tissue culture basal media

Basal medium B5 BDS BABI MS MMS WPM DKW Macronutrient components (mg L−1)

rooting applications. The remaining 6% of the citations were orchid applications using Knudson (1925), VW (Vacin and Went 1949), or chemically non-defined media.

Murashige and Skoog medium is widely used for both dicots and monocots, and it makes a good plant regeneration medium, because of the high levels of nitrogen in both nitrate and ammonium forms, with a relatively high ratio of ammo- nium to nitrate (Table 1). However, sometimes MS medium is not the best medium for optimal growth, precisely because it does contain high levels of ammonium ions (Gamborg et al. 1976). In contrast, B5 medium may be too low in ammonium (Table 1) for some plant species. Dunstan and Short (1977) found deficiencies in B5 medium for Allium cepa L. (onion) tissue culture, and modified B5 by adding a modest amount of

ammonium and phosphate and called the medium BDS (B5 as modified by Dunstan and Short; Table 1). The BDS formula- tion recently was modified with the addition of more calcium, and this version, called BABI (Table 1), which works well with a wide variety of plant species and tissue culture appli- cations, including Nicotiana tabacum L. (tobacco), other di- cots, monocots, and some woody plants (Greenway et al. 2012). Although the extra calcium in BABI does not have a strong effect on biomass growth compared with BDS, in sev- eral cases, an advantage in plant regeneration productivity was reported with the additional calcium (Greenway et al. 2012).

Woody plant basal media generally contain lesser amounts of the macronutrient salts, so ½-MS, MMS (Modified MS), and WPM are now widely used for woody plants. With only half of the MS ammonium, the ½-MS or MMS media do not appear to induce ammonium toxicity. Similarly, WPM has less total nitrogen and less ammonium than MS. In contrast, DKW has a higher ammonium to nitrate ratio similar to MS, but less total nitrogen than MS and it uses different salt sources, resulting in an increase in sulfate concentration (Table 1). Some woody plants may benefit from the greater sulfate in the medium.

There are other basal medium formulations that have been developed for specialty applications. For example, the Nitsch and Nitsch (1969) medium was optimized for tobacco anther cultures to produce haploid plants, and this basal medium has served as a good starting point for androgenesis of other plants. It has approximately half as much total nitrogen as MS, but with a similar ammonium to nitrate ratio. The reduc- tion in nitrogen favors induction of androgenesis (Ferrie and Caswell 2011). Similarly, the NLN (Nitsch and Nitsch as modified by Lichter) basal medium described by Lichter (1982) was developed for isolated microspore culture of Brassica napus L. and contains very low total nitrogen with no ammonium salts. Another example of a specialty basal medium is the Kao and Michayluk (1975) medium developed for protoplast culture, which has been a useful starting point for protoplast cultures of many plant species. This medium contains the long version of the B5 organic supplements (vi- tamins and amino acids), plus additional vitamins, organic acids, sugars, and sugar alcohols. The Kao and Michayluk (1975) medium, rich in organic supplements, provides the numerous metabolic precursors that allow protoplasts to re- generate their cell walls and initiate cell divisions.

In summary, unless a customized system has been devised for a specific application and species, MS or BABI basal me- dia may be recommended for most herbaceous plant tissue culture applications. Alternatives, such as MMS, WPM, or DKW basal media, may be recommended for woody plants.

Incubation conditions Most plant tissue cultures are incubated in reach-in growth chambers or growth rooms. Growth facil- ities should provide temperature control, variable light

intensity of appropriate wavelengths, photoperiod control, and in some cases humidity control (Gamborg and Phillips 1995). Temperatures should be controlled to ± 1°C. Temperature typ- ically remains constant, most frequently in the range of 22– 28°C, depending on the plant species. However, specialty sit- uations may call for short-term (days to weeks) cold or heat treatments, with a lower limit of 4°C and an upper limit of 35°C being most common.

Typically, light intensity is set between 50 and 100 μmol m⁻² s⁻¹ photosynthetically active radiation, but photosynthetic photon flux as low as 5, or as high as 150 μmol m⁻² s⁻¹, have been reported depending on the spe- cies and application. Photoperiod is often set to 16 h, resulting in 8 h dark periods per d, but may be adjusted for the species of interest. Some protocols identify longer day-lengths, up to continuous light incubation. Humidity controls may be neces- sary in extremely dry environments to reduce evaporation of the medium during incubation. Ventilation of culture vessels, culture densities, and other fundamental considerations are discussed in Gamborg and Phillips (1995).

Plant tissue cultures are subcultured to fresh medium on a regular basis, depending on their growth rates (Gamborg and Phillips 1995). Callus cultures, shoot proliferation cultures, and rooting cultures on semi-solid media are typically subcultured at 3–4 wk intervals, but subculture frequency may be as often as weekly or bimonthly, or as seldom as every 8 wk, depending on the culture system and incubation condi- tions. Cell suspensions in batch cultures may be subcultured to fresh medium at biweekly, weekly, or bimonthly intervals, or even continuously, using certain bioreactors. Most plant tissue culture scientists attempt to time the subculture interval to coincide with mid-exponential or early-linear growth phases, in order to maximize growth potential, or time subcultures to preclude the growth-limiting depletion of any nutrient in the basal medium.

Plant Growth Regulators

Synthetic vs. natural Once the basal nutrient demands of plant tissues are met, further developmental responses are stimulat- ed through the addition of growth regulators such as auxins, cytokinins, gibberellins, ethylene (or, more accurately, anti- ethylene), or abscisic acid. For the majority of applications, auxins and cytokinins are the most important of these growth regulators. However, synthetic growth regulators, especially synthetic auxins, such as 2,4-dichlorophenoxyacetic acid (2,4- D) and picloram (PIC), are more potent by perhaps 10–1000 times than natural auxins, such as indole-3-acetic acid (IAA); therefore, amounts used are relative to potency. Auxin potency also varies among plant species. For these reasons, it is im- possible for classical physiologists to compare the action of different natural and synthetic auxins on an equimolar basis,

because a nanomolar amount of PIC may have similar activity to a micromolar amount of IAA. It is therefore common to report growth regulator concentrations in mg L⁻¹, the actual amount of the specific chemical used during media prepara- tion, rather than molarity. Commonly used plant growth reg- ulators, as surveyed from Herman (2015), are presented in Table 2 and discussed in more detail below.

Auxins The 5-y set of media citations (Herman 2015) indicat- ed that α-naphthaleneacetic acid (NAA; 15% of citations) and indole-3-butyric acid (IBA; 9% of citations) are the most fre- quently used auxins, followed by 2,4-D, IAA, and PIC. Natural auxin, IAA, is light sensitive and degrades easily, while the other auxins mentioned are synthetic and chemically more stable. Commonly, NAA, IBA, and IAA are used in shoot multiplication cultures and in rooting cultures. Dicamba (DCA), 2,4-D, and PIC are auxinic herbicides and tend to exhibit activity at lower concentrations than do IAA, IBA, and NAA. Any of these auxins can be used to stimulate callus or cell proliferation. For induction of somatic embryo- genesis, 2,4-D has been used in many species; however, in some species, DCA and PIC can also be used to induce so- matic embryogenesis. Other auxinic herbicides, such as ρ– chlorophenoxyacetic acid and 2,4,5-trichlorophenoxyacetic acid have been employed as well, but as they do not seem to have any unique properties compared to 2,4-D, they are rarely used in current protocols. Following induction, 2,4-D is known to suppress somatic embryo development, which leads to a decline in regeneration capacity over time. Dunstan and Short (1978) reported that regeneration from onion callus was only possible for callus up to 6 mo of age, when 2,4-D was used for induction, whereas, Phillips and Hubstenberger (1987) demonstrated that PIC sustains regeneration capacity over longer periods of time by regenerating plants from 20- mo-old onion callus. Additionally, brassinosteroids have been evaluated as alternative auxins or auxin stimulators, along with several other auxin variants and inhibitors available for specialty situations, and new generation auxinic herbicides are continuously under development (George et al. 2008).

Cytokinins The 5-y set of media citations (Herman 2015) indi- cated that N⁶-benzyladenine (BA) is the most frequently used cytokinin (31% of citations), followed by kinetin (KIN; 7.5% of citations). Thidiazuron (TDZ) and 2-isopentenyladenine (2-IP) were cited less frequently (4% of citations each). Natural cyto- kinins such as zeatin (ZEA) are expensive and less chemically stable than synthetic versions, such as BA. Thidiazuron is more frequently used in woody plant applications (Huetteman and Preece 1993); it appears to have properties in common with both auxins and cytokinins and was originally developed as an herbicide by virtue of its inhibition of cytokinin oxidase. Adenine hemisulfate (ADE) is used as a cytokinin precursor and has weak cytokinin activity. Manipulation of auxin and

Table 2. Phytohormones and growth regulators commonly used in plant tissue culture

cytokinin alone was sufficient to accomplish culture goals in 89% of the cases cited by Herman (2015).

Gibberellin, abscisic acid, ethylene, and others Gibberellic acid (GA₃) was cited in 4% of the cases summarized by Herman (2015) and is commonly used in shoot elongation protocols, but it can also influence seed development and overcome dormancy. Anti-gibberellins, such as paclobutrazol (PBZ) and uniconazole (UNI) are available; they are used to prevent GA₃ synthesis and thus reduce stem elongation. Interestingly, PBZ and UNI were reported in the Herman (2015) citations for use in stimulating rooting of woody plants. Abscisic acid (ABA), commonly used to regulate so- matic embryo development, was employed in 2% of the cases referenced by Herman (2015). Silver nitrate or silver thiosul- fate was used in 5% of the cases in the Herman 5-y survey (2015); silver compounds are frequently used to inhibit the action of ethylene. The synthesis of ethylene can be promoted by 1-aminocyclopropane-1-carboxylic acid (ACC) or 2- chloroethylphosphonic acid (ethephon) (Biddington 1992). Jasmonic acid (JA) is often used as an elicitor in hairy root cultures. While not phytohormones, ascorbic acid (ASA) or polyvinylpyrrolidone (PVP) have been used as antioxidants in 4% of the Herman (2015) 5-y references.

Phytohormone receptors and differential activity Plant tissue culture scientists may spend significant time identifying geno- types that are responsive in culture or determining the explant types that will support the type of growth or developmental response under study. Many have observed that the same growth regulator treatment applied to different explants (tissue types) of the same species or genotype may result in different responses, suggesting possible tissue specificity of phytohor- mone receptors or effectors (Phillips 1988) or the interaction of endogenous phytohormones within the tissues and the ex- ogenously supplied growth regulators in tissue culture.

Phytohormones achieve their effects through signal trans- duction and transcriptional activation mediated initially by binding to specific receptors. At one time, it was thought that there would be only a single receptor per phytohormone class, e.g., only one auxin-binding receptor. However, recent research has identified multiple receptors per phytohormone class. For instance, there has been a 23-member family of auxin response factor (ARF) proteins, a 29-member family of Aux/IAA co- receptor proteins, and a 5-member family of auxin signaling F-Box proteins identified to be associated with auxin activity (Mockaitis and Estelle 2008). Similarly, three cytokinin recep- tors (histidine kinases) localized to the endoplasmic reticulum (ER) have been identified (Wulfetange et al. 2011). Gibberellin

signaling involves two F-Box receptors (Daviere and Achard 2013). Abscisic acid (ABA) receptors include 14 members of the Pyrabactin Resistance (PYR)/PYR-Like (PYL)/Regulatory Components of ABA Receptors (RCAR) family (Kline et al. 2010). There are five structurally and functionally distinct ER- localized receptors for ethylene (Gallie 2015). The last of these authors asked: BWhy so much complexity?^

In contrast to plants, animal hormone models tend to be

discretely compartmentalized in organs, with little direct inter- action among them; thus, the answer to the query above might simply be: there is so much complexity among phytohormone receptors, because phytohormones initiate or mediate so many different physiological processes and interact with each other in various ways during these processes (illustrated in Table 3). A single receptor per phytohormone would seem unable to medi- ate such different responses; whilea single phytohormone bind- ing to a single receptor may have a particular allosteric result and physiological effect, it is difficult to imagine how that would lead to multiple responses. Table 3 illustrates that differ- ent physiological responses do occur, at least in part, because of phytohormone interactions. Another illustration of phytohor- mone interactions is the fact that cytokinin is synthesized in root apices and then transported to the shoot, whereas auxin is synthesized in the shoot apices and transported to roots, thus creating reciprocal gradients throughout the plant tissues.

Another way in which a single phytohormone can affect multiple responses is by subcellular compartmentation (Spartz and Gray 2008); abscisic acid binds to distinct receptors in the plasmalemma, in the nucleus, and in the chloroplast, which may explain some of the different physiological functions of the phytohormone. Similarly, auxin is known to bind to distinct receptors in the cytoplasm and in the nucleus (Mockaitis and Estelle 2008). Compartmentation readily explains how a single phytohormone can have more than one physiological effect. While animal hormones and their receptors tend to be compart- mentalized discretely by organ or tissue type, phytohormone activity appears to be compartmentalized by subcellularly lo- cated receptors with distinct signal transduction pathways.

In addition, auxin receptor–ligand interaction studies have shown that electrostatic differences in binding of IAA vs. PIC lead to differential binding affinities (Calderon Villalobos et al. 2012). This explains why a nanomolar amount of PIC may have similar effect as a micromolar amount of IAA. With multiple receptors and co-receptors for auxin, and differential binding affinities for the various versions of synthetic growth regulators and the phytohormone itself, it can easily be envisioned that different combinations of these components lead to multiple physiological responses. At any given time during development, some of the co-receptors and effectors available vary depending upon tissue and environmental cues. If plant growth and development is based, at least in part, on the allosteric interaction of each phytohormone (or its synthet- ic counterparts) with its subcellular-specific compartment or situation-specific (environmentally cued) receptors and/or co- receptors/effectors, then these differential interactions may ex- plain the distinct effects of growth regulators observed both in vivo and in vitro. Perhaps such differential allosteric/ electrostatic binding effects might explain why 2,4-D used as an auxin seems to have the almost unique ability to induce somatic embryogenesis in tissue cultures of many plant species.

Classical Developmental Models

Tobacco organogenesis The classic demonstration of plant tissue response to growth regulators was reported by Skoog and Miller (1957), using tobacco as a model system to study the effects of IAA and KIN (see Fig. 1). Using a factorial experiment, Skoog and Miller (1957) noticed that shoots formed when cytokinin was high and auxin low, whereas roots formed and proliferated when auxin was moderate to high and cytokinin was absent. Callus (undifferentiated mass of cells) proliferated when moderate amounts of cytokinin were com- bined with high auxin (Fig. 1).

Table 3. Examples of plant developmental processes involving six classes of phytohormones (data extracted from Curaba et al. 2014)

Plant developmental stage

Phytohormone involvement Auxin Cytokinin Abscisic

acid

Gibberellic acid

Ethylene Jasmonic

acid

Embryo Yes Yes No Yes Yes No Germination Yes Yes Yes Yes Yes No Stress response Yes Yes Yes Yes Yes Yes Root Yes Yes Yes Yes Yes Yes

Leaf Yes Yes No Yes Yes Yes Phase transition No No Yes Yes No No Flower Yes Yes No Yes Yes Yes Senescence Yes Yes Yes Yes Yes Yes

Figure 1. Schematic illustration of the Skoog and Miller (1957) Nicotiana tabacum L. (tobacco) phytohormone model for organogenesis. Auxins and cytokinins used in combination elicit an array of tissue responses based on their relative concentrations. Higher relative auxin concentrations increase root formation, whereas higher relative cytokinin concentrations induce shoot formation. When both auxin and cytokinin concentrations are moderate to high, callus is induced.

Increasing cytokinin concentration

Increasing auxin concentration

These observations formed the basis of shoot and root or- ganogenesis phytohormone models and of biomass (callus) growth in tissue culture, and essentially established the mod- ern concept of plant growth regulators. Induction and devel- opment of the organ can occur on the same medium. It is noteworthy that shoots are induced and developed on one medium treatment, while roots are induced and developed on another medium treatment (within typical subculture inter- vals), so that at least two culture steps are required to obtain a whole plant: at least one induction/development treatment for each organ (shoot, root). Some plants are induced for rooting on an auxin-containing medium, then transferred to growth regulator-free medium for development (Gamborg and Phillips 1995), thereby adding an extra culture step.

Daucus carota L. (carrot) somatic embryogenesis Not long after Skoog and Miller’s (1957) discovery of shoot and root formation in tobacco, Reinert (1958) and Steward et al. (1958) observed carrot cells in culture develop into embryos rather than simple organs (Fig. 2). These so-called somatic embryos follow the characteristic developmental stages of zygotic em- bryos. Somatic embryos are induced by high auxin, and 2,4-D seems to have an almost unique capacity to stimulate somatic embryogenesis in many plant species. In some cases, direct somatic embryogenesis occurs without additional cycles of embryogenesis, whereas in other cases, the somatic embryos will proliferate additional series of somatic embryos in an adventitious manner (arising from an unusual location other

than a pre-existing meristem). In yet other species, an embryo- genic callus proliferates, which is then manipulated to recover mature somatic embryos. Developing somatic embryos are bipolar structures, containing both shoot- and root-meristems. Typically, high auxin (usually 2,4-D) induction is followed by phytohormone-free development and maturation of the somat- ic embryos, so at least two culture steps are required to obtain the whole plant: one induction step for the bipolar embryo, and at least one development (maturation) step.

Origins and modes of development The easiest and most di- rect mode to regenerate plants from tissue cultures is by en- hanced axillary branching, which is the basis of most micropropagation or cloning protocols (Fig. 3A). Axillary buds are preformed meristems, and stimulation of axillary branching typically results in a 10-fold increase in shoot num- ber per monthly culture passage, with as many as 1000,000 shoots in 6 mo proliferating from a single explant. A propor- tion of these shoots can be separated and rooted for commerce, while others are used to maintain the shoot proliferation stock. Proliferation of axillary branches or nodes tends to minimize spontaneous mutations in the stock material (Chu 1992).

As illustrated above in the classical developmental models, plants also can be regenerated via organogenesis (Fig. 3B, C) or through somatic embryogenesis (Fig. 3C, D; Gamborg and Phillips 1995). Either of these modes can occur from an ad- ventitious origin (Fig. 3B, D) or from a de novo origin (Fig. 3C). Adventitious and de novo organogenesis and de

Figure 2. Schematic illustration of the Reinert (1958) and Steward et al. (1958) Daucus carota L. (carrot) phytohormone model for somatic em- bryogenesis. The auxin 2,4-dichlorophenoxyacetic acid (2,4-D) is used to proliferate cells and to induce somatic embryogenesis. Upon removal of

2,4-D, the small somatic embryos develop into mature embryos and plantlets, following the same developmental stages as zygotic embryogenesis.

novo somatic embryogenesis systems tend to accumulate more spontaneous mutations, compared to enhanced axillary branching and adventitious somatic embryogenesis (Chu 1992). Adventitious shoots, roots, or somatic embryos arise from organized tissues, but not from preformed meristems (Gamborg and Phillips 1995). Often the presence of vascular or dermal tissues is required for new adventitious structures to form, and adventitious responses are typically shorter-lived, persisting for perhaps 30–60 d of culture. In contrast, de novo

shoots, roots, or somatic embryos Barise anew^ from unorga- nized proliferating tissues such as callus, and de novo re- sponses typically persist for months or even a few years in culture. The de novo systems of plant regeneration are neces- sary for cell selection and other applications requiring longer- term culture manipulation. Adventitious systems can be useful for micropropagation and other shorter-term applications. Either adventitious or de novo systems are used to recover plants following gene transfer, depending upon the species.

Figure 3. Origins and modes of development during in vitro plant regeneration. (A) Mode of axillary branching, originating from preformed meristems. (B) Adventitious origin of the mode of organogenesis, giving rise to shoot organs or root organs occurring at an unusual location on the plant, not from preformed meristems. (C) De

novo origin of regeneration from unorganized callus, illustrated for the mode of organogenesis (shoots, roots) and for the mode of somatic embryogenesis via 2,4-dichlorophenoxyacetic acid (2,4-D). (D) Adventitious origin of the mode of somatic embryogenesis.

Elaborated Developmental Models

Trifolium pratense L. (red clover) A comprehensive system for the culture of cells, tissues, and organs of red clover, Trifolium pratense L., is presented here, because it illustrates a number of principles and concepts that elaborate on the classical developmental models and medium manipulation.

Basal media were evaluated for callus proliferation using five cultivars of red clover (Phillips and Collins 1979a). A series of factorial and split-plot design experiments were per-

Table 4. The nutrient compositions of the Legume version 2 (L2) basal medium (Phillips and Collins 1979a) and its modifications (Phillips and Collins 1979b, 1980) used for tissue culture of Trifolium pratense L. (red clover)

formed using visual ratings of callus proliferation. First, the ^{4 2}

macronutrients were optimized, then the micronutrients were optimized using MS, SH, and B5 as initial control media. More than 60 formulations were assessed to optimize the macro- and micronutrients. More than 20 formulations were compared for carbohydrate and vitamin requirements. More than 45 formulations were compared for growth regulator composition, and these were tested additionally with multiple genotypes of Glycine max (L.) Merr. (soybean) and Medicago sativa L. (alfalfa). This tedious process resulted in the L2 (Legume version 2; Phillips and Collins 1979a) basal medium developed specifically to support growth of forage and grain legumes (Table 4). Biomass fresh weight increases were mea- sured in a final comparison of L2, MS, and SH (B5 was not used because in previous experiments, it had yielded re- sponses identical to those of SH). The L2 basal medium re- sulted in statistically superior callus growth for alfalfa and soybean over MS and SH. Red clover performed equally well on MS or L2, but L2 was superior to SH for callus growth. The Phillips and Collins (1979a) L2 basal medium is similar to MS, but with increased amounts of some nutrients and less ammonium nitrogen. Callus cultures proliferating on high PIC or 2,4-D concentrations regenerated shoots, some of

which spontaneously rooted, when transferred to medium augmented with 0.005 mg L⁻¹ PIC + 1.0 mg L⁻¹ BA.

The red clover tissue culture system was then adapted for virus elimination from cultivar parental clones (Phillips and Collins 1979b). Full-strength L2 basal medium using

0.004 mg L⁻¹ PIC + 1.0 mg L⁻¹ BA appeared to be best for stimulating shoot growth and shoot multiplication by en- hanced axillary branching from explants containing the meri- stematic dome and one primordial leaf (0.1 to 0.4 mm shoot meristem tips). The auxin content had to be slightly reduced in order to eliminate excess callusing. These shoots rooted on half-strength L2 basal medium containing 0.2 mg L⁻¹ IAA (RL; Rooting on modified L2) shown in Table 4. The rooting medium was developed by comparing full-strength, ½- strength, and ¼-strength versions of L2, with ½-strength being best. Also, different auxins were tested before arriving at

0.2 mg L⁻¹ IAA (Phillips and Collins 1979b).

The callus to plant regeneration system was elaborated fur- ther in a subsequent study using cell suspensions of red clover (Phillips and Collins 1980). The basal medium for cell

suspension culture, SL2 (Suspensions in modified L2; Table 4), was adjusted by lowering the amounts of macronu- trient salts (approximately 90% of L2) and especially ammo- nium nitrogen, because full-strength L2 appeared to be bor- derline toxic to liquid cell suspension cultures (Phillips and Collins 1979a). By taking cultures through a cycle from fine cell suspension back to callus and using 0.06 mg L⁻¹ PIC +

0.1 mg L⁻¹ BA for both steps, it was apparent that somatic

embryogenesis was the mode of plant regeneration (Phillips and Collins 1980). That is, cotyledons emerged, followed by the appearance of a unifoliate leaf, which in turn was followed by the development of trifoliate leaves—the typical legume embryo–to–plant developmental stages. Although a range of concentrations of PIC or 2,4-D supported the induction of somatic embryogenesis, Phillips and Collins (1980) deter- mined that the best induction of somatic embryos occurred using 0.01 mg L⁻¹ 2,4-D + 2 mg L⁻¹ ADE. Subsequently, somatic embryos were transferred to shoot pole promotion medium containing 0.001 mg L⁻¹ PIC + 0.2 mg L⁻¹ BA and were eventually rooted as before. Later this strategy was

extended to regeneration of plants from clover protoplasts (Grosser and Collins 1984; Myers et al. 1989).

The tissue culture system was then adapted to rescue im- mature heart-staged hybrid zygotic embryos from interspecif- ic crosses (Phillips et al. 1982). Immature zygotic embryos were rescued and developed initially using L2 basal medium + 12.5% (w/v) sucrose (to inhibit precocious root germina- tion) + 0.006 mg L⁻¹ PIC + 2 mg L⁻¹ ADE, designated LIH (L2 for Interspecific Hybrid embryos). These Bmatured^ em- bryos were transferred to a version of the shoot pole promo- tion medium (as used with somatic embryos), the best version in this case being 0.001 mg L⁻¹ PIC + 0.15 mg L⁻¹ BA. The BA concentration had to be adjusted downward from

0.2 mg L⁻¹ to eliminate excess callusing. Once shoots emerged on this medium, they were transferred to a version of the shoot multiplication medium (as used with shoot mer- istem tips), the best version in this case being 0.003 mg L⁻¹ PIC + 0.5 mg L⁻¹ BA. Shoots were rooted as before. This strategy also worked for a second interspecific hybrid (Phillips et al. 1992).

As can be observed from the history above, some develop- mental steps were optimized in an evolving manner as the system was applied to other situations or explants. All of these developmental steps were tied together into a model system (Collins and Phillips 1982; Phillips and Collins 1984), illus- trated in Fig. 4. This model (Fig. 4) identifies discrete devel- opmental stages and their relationships: Seedlings may be germinated under sterile conditions on SGL (Seed Germination modified L2; Table 4) medium to provide ex- plants. Explants produce callus on L2 medium. Callus may be used to establish cell suspension cultures using SL2 medi- um, and suspensions may be cycled back to callus (using L2 medium), prior to regeneration. Also, protoplasts from cell suspensions or from leaf explants are recovered as cell sus- pensions and cycled to callus for regeneration. Callus from

any of these sources may be induced for somatic embryogen- esis on LSE (L2 for Somatic Embryogenesis) medium. Somatic embryos are transferred to LSP2 (L2 for Shoot Pole promotion v.2) medium for embryo shoot pole promotion and germination. Immature hybrid heart-staged zygotic embryos are matured on LIH medium, and then the shoot poles are promoted using LSP2 medium. Whether derived from somat- ic embryos or immature zygotic embryos, shoots are multi- plied by enhanced axillary branching using ML8 (Multiplication on L2 v.8) medium. Shoot tips of any clone of red clover or shoot meristem tips as used for virus elimina- tion also proliferate multiple shoots using ML8 medium. Regardless of the source of shoots, they are rooted using RL medium. The resulting plantlets are ready for acclimatization in a growth chamber or greenhouse, and eventual transfer into a field for further study.

The red clover model illustrates two distinct ways in which the medium can be manipulated to achieve specific develop- mental or growth responses. First, the basal medium can be adjusted for specific developmental steps (Table 4), and, sec- ond, the growth regulators can be adjusted for specific devel- opmental steps (Table 5).

The L2 Basal medium is the full-strength version used for tissue and organ proliferation (Table 4). The SL2 medium has less ammonium nitrogen and a little less of the other macro- nutrient salts to support growth of cell suspensions, whereas the RL medium is a half-strength version of L2 used specifi- cally for rooting of shoots (Table 4). Seed Germination mod- ified L2 (SGL) medium, used for in vitro seed germination, is approximately a 1/10-strength version of L2 major salts with no micronutrient salts, but with ¼-strength vitamins and 40% of the sucrose (Table 4).

Most of the developmental manipulation is accomplished through the use of growth regulators (Table 5): Seed germina- tion on SGL does not require any growth regulators. Callus or

Figure 4. Discrete developmental steps in Trifolium pratense L. (red clover) tissue culture (adapted from Phillips and Collins 1984). LSE, L2 modified for somatic embryogenesis; LSP2, L2 for shoot pole promotion version 2; LIH, L2 modified for interspecific hybrid embryos; SL2, suspension culture modified L2; ML8, multiplication modified L2 version 8; L2, legume medium version 2; RL, rooting modified L2; SGL, seed germination modified L2.

‘ML8’

Table 5. Growth regulator manipulations used in tissue culture of Trifolium pratense L. (red clover) adapted from Phillips and Collins (1984)

Medium designation

Use Basal medium

(from Table 3)

Auxin Cytokinin

2,4-D, 2,4-dichlorophenoxyacetic acid; PIC, picloram (4-amino-3,5,6-trichloropicolinic acid); IAA, indole-3-acetic acid; BA, N6-benzyladenine (6- benzylaminopurine); ADE, adenine hemisulfate

cell suspension proliferation on L2 or SL2 medium involves a relatively high level (0.06 mg L⁻¹) of the auxin PIC along with a moderately low level (0.1 mg L⁻¹) of BA as cytokinin. Somatic embryogenesis is induced on LSE medium using a very low level of (0.01 mg L⁻¹) the auxin 2,4-D, plus very weak cytokinin activity using ADE (2 mg L⁻¹; Table 5). Heart-staged hybrid embryos are cultured on LIH medium with 12.5% (w/v) sucrose to prevent precocious germination of the root pole, supplemented with a low level (0.006 mg L⁻¹) of the auxin PIC, plus very weak cytokinin activity, using 2 mg L⁻¹ ADE (Table 5). Both somatic embryos and immature zygotic embry- os are encouraged to express the shoot pole using a very low level (0.001 mg L⁻¹) of PIC, plus a moderate level (0.15 mg L⁻¹) of BA in LSP2 medium. Shoots are multiplied on ML8 medium using a moderately low level (0.003 mg L⁻¹) of PIC as auxin plus a high level (0.5 mg L⁻¹) of BA as cyto- kinin. Shoots are rooted on RL medium using a moderate auxin level of 0.2 mg L⁻¹ IAA (Table 5).

This is one optimized model or approach; there are other variations that are equally effective in other systems. For example, in the Finer–Parrott soybean system (Finer and Nagasawa 1988; Samoylov et al. 1998), somatic em- bryos are induced from primary explants, and then those primary somatic embryos are proliferated in large num- bers in suspension culture. Individual somatic embryos are matured and germinated into plants. Both the clover and soybean methods include a cloning step: shoots from somatic embryos are cloned in the red clover approach, while somatic embryos themselves are cloned in the soy- bean approach.

There are intriguing parallels between induction of somatic embryogenesis and development of heart- staged hybrid embryos in the red clover system: both need a low level of auxin (though a different source in each case) and a weak cytokinin signal. Also, the trend

of auxin and cytokinin signals from induction through expression shows a pattern: induction involves low aux- in + weak cytokinin; then, shoot pole promotion re- quires a very low auxin + moderately low cytokinin; finally, shoot multiplication benefits from a moderate level of auxin + high cytokinin. The cytokinin signal starts low and increases with each step to a higher lev- el, while the auxin signal fluctuates with each step. It is known that auxin plays a critical role in shoot apical organization (Wang and Jiao 2018); thus, these auxin fluctuations from induction through development might be interpreted as likely to be related to apical organiza- tional steps.

Extension to onion and Oryza sativa L. (rice) Although the clover model was developed for a dicot, elements of the clover model were tested on the monocot Allium spp., (onion). Basal media were screened and BDS (Dunstan and Short 1977) was found to be superior to others for onion tissue culture growth (Phillips and Luteyn 1983). Callus production was better using 2.0 mg L⁻¹ BA + 0.75 mg L⁻¹ PIC compared to NAA or 2,4-D as auxin, and this combination also induced somatic embryogenesis. This callus developed somatic embryos when transferred to 0.03 mg L⁻¹ PIC + 0.35 mg L⁻¹ BA (Phillips and Luteyn 1983; Phillips and Hubstenberger 1987). As so- matic embryos developed their shoot poles, they were trans- ferred to 0.03 mg L⁻¹ PIC + 0.5 mg L⁻¹ BA for shoot multi- plication. This regeneration system also applied to onion pro- toplasts (Hansen et al. 1995). In that case, isolated protoplasts from cell suspensions initiated on either 2,4-D- or PIC-based medium were maintained on 2.0 mg L⁻¹ NAA + 0.5 mg L⁻¹ BA. Cell colonies and calluses were recovered from proto- plasts using 1.0 mg L⁻¹ NAA + 0.5 mg L⁻¹ ZEA. These cal- luses were transferred to 0.03 mg L⁻¹ PIC + 0.35 mg L⁻¹ BA for development of somatic embryos.

Subsequently, Oryza sativa L. (rice) was tested in a manner similar to onion. Once again, BDS was found to be superior to MS or N6 (Chu 1978) as the basal medium (Chowdhury et al. 2006). Callus cultures were started on 2.2 mg L⁻¹ 2,4-D or

5.0 mg L⁻¹ PIC and were transferred to 0.03 mg L⁻¹ PIC +

0.35 mg L⁻¹ BA for development of somatic embryos and shoot poles (Chowdhury et al. 2006). The resulting rapid rice regener- ation system used the same shoot pole promotion medium as was used in onion, with similar success (Dabul et al. 2009).

Onion and rice, both monocots, seem to be about 10× less sensitive to PIC, compared to red clover, a dicot plant. Callus formation in onion used 0.75 mg L⁻¹ PIC (Phillips and Luteyn 1983) vs. 0.06 mg L⁻¹ PIC for red clover (Phillips and Collins 1980). Clover shoot pole pro- motion used 0.001 mg L⁻¹ PIC (Phillips et al. 1982), but for onion or rice shoot pole promotion, 0.03 mg L⁻¹ PIC had to be used (Phillips and Hubstenberger 1987; Dabul et al. 2009). Shoot pole promotion in onion and rice also used a higher amount of BA, 0.35 mg L⁻¹ (Phillips and Hubstenberger 1987; Dabul et al. 2009) vs. 0.15 mg L⁻¹ for clover (Table 5; Phillips and Collins 1984). The results from these contrasting plant systems lead to the sugges- tion that a range of shoot pole promotion media using

0.001 to 0.03 mg L⁻¹ PIC + 0.15 to 0.35 mg L⁻¹ BA

might be used to screen other species.

Pinus brutia var. eldarica (Medw.) Silba (Eldarica pine) Differences in tissue culture requirements are not limited to dicots and monocots; gymnosperms have their special stipu- lations as well. Pinus brutia var. eldarica (Medw.) Silba (Eldarica pine) presents an interesting case study. Basal media were screened, and it was found that MMS (Table 1), with half-strength macronutrient salts and full-strength

micronutrient salts, was most suitable (Gladfelter and Phillips 1987). A de novo organogenesis regeneration system from up to 3-yr-old callus, maintained on medium containing

0.5 to 1.0 mg L⁻¹ NAA + 0.5 to 1.0 mg L⁻¹ BA (Fig. 5;

Gladfelter and Phillips 1987), was developed. The interesting aspect of this regeneration system was that four distinct cul- ture manipulations were required to obtain shoots (Fig. 5): The first step was bud induction, using 0.05 mg L⁻¹ IBA +

1.0 mg L⁻¹ KIN. The second step was a resting phase on

growth regulator-free MMS medium, needed for bud matura- tion. Callus could be recycled between these two media to enhance competence for regeneration. Once the buds had ma- tured sufficiently on MMS, the third step was apical organi- zation, including early needle development (shoot pole pro- motion), using 0.05 mg L⁻¹ IBA + 0.1 mg L⁻¹ KIN. These buds were advanced to step four, shoot elongation using growth regulator-free MMS. Some shoots were rooted using a 1-d pulse of a high level (5.0 mg L⁻¹) of NAA. Once again, a low level of auxin (0.05 mg L⁻¹ IBA) and a moderate level of cytokinin (0.1 mg L⁻¹ KIN) encouraged shoot apical organi- zation in this de novo regeneration system. The same medium as used for bud induction, MMS + 0.05 mg L⁻¹ IBA +

1.0 mg L⁻¹ KIN, was used successfully for short-term adven- titious shooting from explants and for enhanced axillary branching of Eldarica pine, as well as for de novo shoot or- ganogenesis (Fig. 5).

This Eldarica pine model developed by Gladfelter and Phillips (1987) demonstrates that cycling between defined media may enhance acquisition of competence for regenera- tion, as well as shows that multiple steps of growth regulator manipulation may be needed to achieve individual organ or plant recovery. These concepts were also being developed along independent lines of evidence by Christianson (1987)

Figure 5. Discrete developmental steps in Pinus brutia var. eldarica (Medw.) Silba (Eldarica pine) de novo organogenesis. Growth regulators shown in mg L⁻¹ (Gladfelter and Phillips 1987, reproduced with permission from Springer Nature, Switzerland AG). MMS, modified Murashige and Skoog medium; NAA, α– naphthaleneacetic acid; BA, N⁶– benzyladenine; IBA, indole-3- butyric acid.

using Convolvulus arvensis L. (field bindweed). Christianson (1987) demonstrated that pluripotency (i.e., capability of mul- tiple organogenic responses, such as shoot vs. root vs. callus) was a short-term phenomenon, that cycling between defined media may enhance acquisition of competence for a particular organogenic response, and that the regeneration process can be arrested at several distinct points of time by different treat- ments. Both the Eldarica pine and the Convolvulus system appear to be consistent with each other in these respects, es- pecially if the assumption is made that the multiple steps of growth regulator manipulation observed in Eldarica pine cul- tures correspond to overcoming points of organogenic arrest in Christianson’s model.

There is also a parallel between the Eldarica pine multi-step regeneration process and the soybean somatic embryogenesis system developed by the aforementioned Finer (Finer and Nagasawa 1988) and Parrott (Samoylov et al. 1998) labs. The soybean induction medium was optimized primarily by increasing the ammonium to nitrate ratio and total nitrogen amounts, along with high 2,4-D (Finer and Nagasawa 1988). Once induced, somatic embryos were proliferated using lower levels of 2,4-D and lower total nitrogen. Histo-differentiation and maturation of these somatic embryos into whole plants required another two culture treatments, involving the use of maltose as a carbohydrate source and activated charcoal, followed by removal of the activated charcoal (Samoylov

Figure 6. Photomicrography of callus development and de novo shoot organogenesis in Pinus brutia var. eldarica (Medw.) Silba (Eldarica pine). (A) Longitudinal section of cotyledon explant showing vascular organization

(V) and dermal layers (D). (B) Callus proliferation after 2 mo, exhibiting remnants of the vascular trace (V). (C) Callus

2 mo of age illustrating random location of a file (F) of specialized

cells. (D) Unorganized callus after 6 mo of culture. (E) Early stage of de novo bud induction in callus 18 mo of age; note the subsurface concentration of dividing cells (D), surrounded by moribund cells (M). (F) Note nonrandom planes of cell division and vascular development (V) during de novo bud induction. (G) Evidence of dermal layer formations (D) deep within the callus undergoing de novo bud

induction. (H) Advanced stages

of apical organization (A) and early stages of needle development (N) in de novo shoot buds from 2-yr-old callus (Wagley et al. 1987).

et al. 1998). An alternative method involved a histo- differentiation and maturation medium followed by desicca- tion of somatic embryos, prior to germination into plants. Thus, recalcitrant species may need multiple steps to induce regeneration and plant recovery.

The companion paper on Eldarica pine by Wagley et al. (1987) showed the correlation of external culture features with internal (histological) features (Fig. 6). Axillary branching was maintained for up to 4 mo in culture, and adventitious shooting from explants occurred for up to 5 mo, then both responses disappeared. Callus production occurred during the initial 3 mo of culture (Fig. 6B), with a gradual disappear- ance of vascular tissues and specialized cells (Fig. 6A), and proliferation of non-specialized meristematic clusters of cells (Fig. 6C). Unorganized callus (Fig. 6D) was predominant in 3 to 6-mo-old cultures. Cultures 6 to 20 mo old were treated with the bud induction medium (Fig. 6E). The first visible

internal change following bud induction was that the most actively dividing cells were associated with layers of dark moribund (denoted by (M); Fig. 6E) cells. The next visible internal change was the formation of specialized cells, such as vascular cells, and organization of vascular bundles (Fig. 6F). Then, the formation of dermal layers of cells (Fig. 6G) and primordial needles (Fig. 6H) was observed. By this time, pri- mordial buds were evident in the external features of the cal- lus. As the cultures were cycled through the other regeneration media, needles became better formed and shoot buds with clear apical structure emerged on the surface of the cultures. This study by Wagley et al. (1987) illustrates how dedifferen- tiation into callus occurs, and how it is then followed by re- differentiation back into new organized structures.

Capsicum spp. Another plant genus recalcitrant in tissue cul- ture are the chile peppers, Capsicum spp. (Kothari et al. 2010).

Figure 7. Depiction of complex tissue of Capsicum baccatum L. (chile pepper) from half-seed explants containing the radicle and hypocotyl end of the explant, showing adventitious buds transitioning to organogenic callus, with subsequent plant regeneration: (A) 5-wk-old explant with adventitious buds,

1. 4-mo-old organogenic callus,
2. section of a 4-mo-old callus, (arrow 1, soft whitish core; arrow 2, green portion of hard tissue; arrow 3, cluster of buds), and (D) plant regeneration from organogenic callus (Valera- Montero and Phillips 2005, reproduced with permission from the Society for In Vitro Biology).

Even though peppers are in the same family as tobacco, Solanum lycopersicum L. (tomato), and Petunia spp., peppers are considerably more challenging to adapt to tissue culture than other members of the Solanaceae family. Most pepper tissue culture responses have been adventitious and short- term in nature, with few, if any, long-term callus de novo regeneration responses reported. A major limitation seems to be recalcitrance for establishing shoot apical organization and shoot elongation (Hyde and Phillips 1996), which was ad- dressed in the same study with the use of silver nitrate, show- ing modest improvement in successful shoot elongation.

In search of a long-term callus regeneration system for peppers, the half-seed explant of Capsicum baccatum L. was evaluated (Fig. 7A; Valera-Montero and Phillips 2005). Explants were cultured on MS medium + 5.0 mg L⁻¹ BA +

1.0 mg L⁻¹ IAA + 2.0 mg L⁻¹ GA₃, where they proliferated a

mass of semi-organized tissue (Fig. 7B) for up to 3 y, while retaining the ability to regenerate plants. Shoot buds elongated into plantlets (Fig. 7D) when transferred to growth regulator- free medium. These semi-organized tissues were composed of three distinct tissue types shown in Fig. 7C: (1), friable, whit- ish, non-organogenic tissue, which could be eliminated upon transfer; (2), hard green tissue with vascular tissues, few or no buds; and (3), amorphous hard green tissue with many primor- dial shoot buds. This latter type (3) was the organogenic cal- lus, but it could only be maintained in the presence of some of the (2) type tissue. If all (2) type tissue was eliminated, then bud formation or development stopped. Apparently, the vas- cular structures in the (2) type tissue were necessary to sustain the formation of buds observed in the (3) type tissue.

This kind of regeneration system appears to have parallels with the woody plant nodule system reported by McCown et al. (1988) for Populus spp. Both cases (Valera-Montero and Phillips 2005; McCown et al. 1988) involved semi-organized tissues, which could be proliferated for extended periods of time, while retaining the ability to regenerate plants. Because there was not a complete loss of organization in the callus prior to shoot bud induction, and a certain level of tissue organization was necessary to support the proliferation of these structures, it would seem that these systems are a type of extended adventi- tious regeneration rather than a de novo system.

There is a further parallel between the C. baccatum system and the development of the Zea mays L. (maize) somatic em- bryogenesis system (Armstrong and Green 1985). Both of these systems are dependent upon the recognition of specific tissue types in complex cultures in order to isolate proliferat- ing tissues capable of long-term plant regeneration.

Conclusions

The field of plant tissue culture has matured to the point where many scientists are seeking to use well-defined Bcookbook^

approaches to apply tissue culture systems to current studies. For a number of plant species, such cookbook recipes are available and can be applied successfully, without adaptation or special training. However, there are still many species where little tissue culture research has been conducted and available recipes may or may not be optimal. It does not make sense to customize the medium for a given species or situa- tion, unless the correct starting point is identified (e.g., starting with MS vs. BABI vs. MMS vs. WPM vs. DKW), nor does it make sense to conduct high-throughput phenotyping using tissue culture systems, when non-optimized protocols are used for the given species or situation. The basic principles of how to manipulate the basal medium and plant growth regulators to achieve better-optimized tissue cultures in little-researched plant species have been outlined in this present review. These principles assume that responsive genotypes and re- sponsive explants can be identified within the given species, and that the operator can recognize tissue organizational pro- gression through developmental steps.

Acknowledgments The authors thank Dr. John Finer for his encourage- ment and guidance for this article. The authors also thank the reviewers for their helpful comments.

References

Armstrong CL, Green CE (1985) Establishment and maintenance of fri- able, embryogenic maize callus and the involvement of L-proline. Planta 164:207–214

Biddington NL (1992) The influence of ethylene in plant tissue culture.

Plant Growth Regul 11:173–187

Calderon Villalobos LIA, Lee S, Oliveira C, Ivetac A, Brandt W, Armitage L, Sheard LB, Tan X, Parry G, Mao H, Zheng N, Napier R, Kepinski S, Estelle M (2012) A combinatorial TIR1/AFB-Aux/ IAA co-receptor system for differential sensing of auxin. Nat Chem Biol 8:477–485

Chowdhury MRDA, Hubstenberger J, Phillips G (2006) Development of in vitro regeneration protocols for Arkansas rice varieties (Oryza sativa L.). J Arkansas Acad Sci 60:108–112

Christianson ML (1987) Causal events in morphogenesis. In: Green CE, Somers DA, Hackett WP, Biesboer DD (eds) Plant tissue and cell culture. Alan R. Liss, New York, NY, pp 45–55

Chu CC (1978) The N6 medium and its applications to anther culture of cereal crops. In: Proc. Symp. On Plant Tissue Culture. Science Press, Beijing, China, pp 43–50

Chu IYE (1992) Perspectives of micropropagation industry. In: Kurata K, Kozai T (eds) Transplant production systems. Kluwer Academic, Amsterdam, The Netherlands, pp 137–150

Collins GB, Phillips GC (1982) In vitro tissue culture and plant regener- ation in Trifolium pratense L. In: Earle ED, Demarly Y (eds) Variability in plants regenerated from tissue cultures. Praeger Scientific, New York, NY, pp 22–34

Curaba J, Singh MB, Bhalla PL (2014) miRNAs in the crosstalk between phytohormone signaling pathways. J Exp Bot 65:1425–1438

Dabul ANG, Belefant-Miller H, Chowdhury MR, Hubstenberger JF, Lorence A, Phillips GC (2009) Screening of a broad range of rice (Oryza sativa L.) germplasm for in vitro rapid plant regeneration and development of an early prediction system. In Vitro Cell Dev Biol– Plant 45:414–420

Daviere JM, Achard P (2013) Gibberellin signaling in plants.

Development 140:1147–1151

Driver JA, Kuniyuki AH (1984) In vitro propagation of Paradox walnut rootstocks. HortScience 19:507

Dunstan DI, Short KC (1977) Improved growth of tissue cultures of the onion, Allium cepa. Physiol Plant 41:70–72

Dunstan DI, Short KC (1978) Shoot production from onion callus tissue cultures. Scientia Hort 9:99–110

Ferrie AMR, Caswell KL (2011) Isolated microspore culture techniques and recent progress for haploid and doubled haploid plant produc- tion. Plant Cell Tissue Organ Cult 104:301–309

Finer JJ, Nagasawa A (1988) Development of an embryogenic suspen- sion culture of soybean (Glycine max Merrill.). Plant Cell Tissue Organ Cult 15:125–136

Gallie DR (2015) Ethylene receptors in plants—why so much complex- ity? F1000Prime Rep 7:39

Gamborg OL, Miller RA, Ojima K (1968) Nutrient requirements of sus- pension culture of soybean root cells. Exp Cell Res 50:151–158

Gamborg OL, Murashige T, Thorpe TA, Vasil IK (1976) Plant tissue culture media. In Vitro 12:473–478

Gamborg OL, Phillips GC (eds) (1995) Plant cell tissue organ culture: fundamental methods. Springer, Heidelberg, Germany, p 358

George EF, Hall MA, de Klerk GJ (2008) Plant growth regulators I: introduction; auxins, their analogues and inhibitors. In: George EF, Hall MA, de Klerk GJ (eds) Plant propagation by tissue culture, 3rd edn. Springer, Heidelberg, Germany, pp 175–204

Gladfelter HJ, Phillips GC (1987) De novo shoot organogenesis of Pinus eldarica Medw. in vitro. I. Reproducible regeneration from long- term callus cultures. Plant Cell Rep 6:163–166

Greenway MB, Phillips IC, Lloyd MN, Hubstenberger JF, Phillips GC (2012) A nutrient medium for diverse applications and tissue growth of plant species in vitro. In Vitro Cell Dev Biol–Plant 48:403–410

Grosser JW, Collins GB (1984) Isolation and culture of Trifolium rubens protoplasts with whole plant regeneration. Plant Sci Lett 37:165–170 Hansen EE, Hubstenberger JF, Phillips GC (1995) Regeneration of shoots from cell suspension-derived protoplasts of Allium cepa L. Plant

Cell Rep 15:8–11

Herman EB (2015) Recent advances in plant tissue culture XXI. Media and techniques for growth, regeneration and storage: 2011–2015. Agritech Consultants Inc., Shrub Oak, NY 148 pp

Huetteman CA, Preece JE (1993) Thidiazurn: a potent cytokinin for woody plant tissue culture. Plant Cell Tissue Organ Cult 33:105–109

Hyde C, Phillips GC (1996) Silver nitrate promotes shoot development and plant regeneration of chile pepper (Capsicum annuum L.) via organogenesis. In Vitro Cell Dev Biol–Plant 32:72–80

Kao KN, Michayluk MR (1975) Nutritional requirements for growth of Vicia hajastana cells and protoplasts at a very low population den- sity in liquid media. Planta 126:105–110

Kline KG, Sussman MR, Jones AM (2010) Abscisic acid receptors. Plant Physiol 154:479–482

Knudson L (1925) Physical study of the symbiotic germination of orchid seeds. Bot Gaz 79:345–379

Kothari SL, Joshi A, Kachhwaha S, Ochoa-Alejo N (2010) Chilli peppers—a review on tissue culture and transgenesis. Biotechnol Adv 28:35–48

Lichter R (1982) Induction of haploid plants from isolated pollen of

Brassica napus. Pflanzenphysiol 105:427–434

Lloyd G, McCown B (1980) Commercially feasible micropropagation of mountain laurel, Kalmia latifolia, by use of shoot tip culture. Comb Proc Int Plant Prop Soc 30:421–427

McCown BH, Zeldin EL, Pinkalla HA, Dedolph RR (1988) Nodule cul- ture: a developmental pathway with high potential for regeneration, automated micropropagation, and plant metabolite production from woody plants. In: Hanover JW, Keathley DE, Wilson CM, Kuny G (eds) Genetic manipulation of woody plants. Basic Life Sciences, vol 44. Springer, Boston, MA, pp 149–166

Mockaitis K, Estelle M (2008) Auxin receptors and plant development: a new signaling paradigm. Annu Rev Cell Dev Biol 24:55–80

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

Myers JR, Grosser JW, Taylor NL, Collins GB (1989) Genotype- dependent whole plant regeneration from protoplasts of red clover (Trifolium pratense L.). Plant Cell Tissue Organ Cult 113:127

Nitsch JP, Nitsch C (1969) Haploid plants from pollen grains. Science 163:85–87

Phillips GC (1988) Developmental models for the expression of totipo- tency and plant regeneration in vitro. In: SAAS Bulletin: Biochemistry and Biotechnology, Vol. 1, Ballal SK (ed), pp. 12–

16. Tennessee Tech. Univ., Cookeville, TN

Phillips GC, Collins GB (1979a) In vitro tissue culture of selected le- gumes and plant regeneration from callus cultures of red clover. Crop Sci 19:59–64

Phillips GC, Collins GB (1979b) Virus symptom-free plants of red clover using meristem culture. Crop Sci 19:213–216

Phillips GC, Collins GB (1980) Somatic embryogenesis from cell sus- pension cultures of red clover. Crop Sci 20:323–326

Phillips GC, Collins GB (1984) Red clover and other forage legumes. In: Evans DA, Sharp WR, Ammirato PV, Yamada Y (eds) Handbook of plant cell culture, vol 2. MacMillan Publ, New York, NY, pp 169–210 Phillips GC, Collins GB, Taylor NL (1982) Interspecific hybridization of red clover (Trifolium pratense L.) with T. sarosiense Hazsl. using

in vitro embryo rescue. Theor Appl Genet 62:17–24

Phillips GC, Grosser JW, Berger S, Taylor NL, Collins GB (1992) Interspecific hybridization between red clover and Trifolium alpestre using in vitro embryo rescue. Crop Sci 32:1113–1115

Phillips GC, Hubstenberger JF (1987) Plant regeneration in vitro of selected

Allium species and interspecific hybrids. HortScience 22:124–125 Phillips GC, Luteyn KJ (1983) Effects of picloram and other auxins on

onion tissue cultures. J Amer Soc Hort Sci 108:948–953

Reinert J (1958) Morphogenese und ihre Kontrolle an Gewebekulturen aus Karotten. Naturwissenschaft 45:344–345

Samoylov VM, Tucker DM, Thibaud-Nissen F, Parrott WA (1998) A liquid-based protocol for rapid regeneration from embryogenic soy- bean cultures. Plant Cell Rep 18:49–54

Schenk RU, Hildebrandt AC (1972) Medium and techniques for induc- tion and growth of monocotyledonous and dicotyledonous plant cell cultures. Can J Bot 50:199–204

Skoog F, Miller CO (1957) Chemical regulation of growth and organ forma- tion in plant tissues cultured in vitro. Soc Exp Biol Symp 11:118–131 Spartz AK, Gray WM (2008) Plant hormone receptors: new perceptions.

Genes Dev 22:2139–2148

Steward FC, Mapes MO, Mears K (1958) Growth and development of cultured cells. II. Organization in cultures grown from freely suspended cells. Am J Bot 45:705–708

Vacin EF, Went FW (1949) Some pH changes in nutrient solutions. Bot Gaz 110:605–613

Valera-Montero L, Phillips GC (2005) Long-lasting Capsicum baccatum ‘organogenetic callus’ formation. In Vitro Cell Dev Biol–Plant 41: 470–476

Wagley LM, Gladfelter HJ, Phillips GC (1987) De novo shoot organo- genesis of Pinus eldarica Medw. in vitro. II. Macro- and micro- photographic evidence of de novo regeneration. Plant Cell Rep 6: 167–171

Wang Y, Jiao Y (2018) Auxin and above-ground meristems. J Exp Bot 69:147–154

Wulfetange K, Lomin SN, Romanov GA, Stolz A, Heyl A, Schmulling T (2011) The cytokinin receptors are located mainly to the endoplas- mic reticulum. Plant Physiol 156:1808–1818