Introduction

After the discovery of cytokinins (CKs) (Hall and De Ropp 1955; Miller et al. 1955; Letham 1963) the science of plant tissue culture (PTC) grew rapidly and exponentially with an equally large expansion in the field of their application (Vasil 2008). In addition to the valuable contributions of PTC to plant propagation, improvement and conservation (Castelblanque et al. 2010; Faizal et al. 2011; Parimalan et al. 2011), it has become a fundamental tool to understand various physiological processes which were once explained using speculative hypotheses and growth parameters such as number of regenerated shoots and roots as well as their survival ex vitro. The ultimate goal of PTC protocols is the rapid multiplication of uniform and healthy plantlets. During micropropagation of many plant species however, physiological disorders such as stunted growth, epigenetic and somaclonal variation are often encountered (Bairu et al. 2011a; Smulders and De Klerk 2011). These problems reduce the commercial application of PTC protocols in several plant species. Therefore, plant growth regulators (PGRs) especially CK, often implicated in the occurrence of these challenges, warrant more stringent studies. Moreover, the choice of CK remain critical to the success or failure of any micropropagation endeavour (Werbrouck 2010; Amoo et al. 2011).

Presently, information on the structure–activity relationship of CKs and its effect on plant physiology is having great significance in understanding the growth requirements of plant species in vitro (Sakakibara 2006). Most plant species exhibit inherent variation which is translated to their varied responses to the different CKs. As a result, it becomes pertinent to optimize PTC protocols for individual species. Researchers are continuously searching for new as well as superior CKs. Topolins in general and meta-topolin (mT) in particular are products of such endeavours. Since the discovery of mT and its derivatives as naturally occurring aromatic CKs in several plant species (Jones et al. 1996; Strnad et al. 1997; Tarkowská et al. 2003), their use in PTC increased rapidly (Table 1). Positive reports on important PTC parameters such as shoot multiplication rate, alleviating physiological disorders, better acclimatization and rooting are making topolins popular amongst plant tissue culturists. Despite the use of topolins for more than a decade, a comprehensive and detailed review on the various research outputs is still lacking. Werbrouck (2010) briefly reviewed the merits and drawbacks of the new aromatic CKs in PTC. The chemistry and synthesis as well as natural occurrence of topolins in plants has also been examined (Subbaraj 2011). This author also reviewed the structure–activity relationships and diverse applications of topolins in agriculture as well as the effect on abiotic stress management. Nevertheless, other important aspects of current research outputs on topolins especially the practical applications remain to be fully exploited. Since the discovery of topolins, our research groups (South Africa and Czech Republic) have tested these compounds on a number of species ranging from food crops, medicinal plants to plantation trees. Here, we compile (Table 1), discuss and summarize the various reports on the use of topolins in PTC when compared to other commonly used CKs.

Table 1 The influence of topolins in comparison to other cytokinins on various growth parameters and physiological disorders
Full size table

The science of PTC in general and CKs in particular are amongst the most reviewed subjects. This review aims at providing background reading for the tissue culture community to help them augment and evaluate the various scientific reports on PTC and the use of topolins. The review therefore, discusses aspects of CK metabolism, structure–activity relationships as well as associated problems of plant growth and development, from a tissue culturist’s perspective.

Historical background

Following the discovery of kinetin (KIN) (Hall and De Ropp 1955; Miller et al. 1955) and later zeatin (Z) (Letham 1963), many other compounds with CK-like activity have been discovered (Schmülling 2004; Sakakibara 2006). At present, there are hundreds of natural and synthetic CKs. Based on their side chain configuration, the naturally occurring CKs are classified as either isoprenoid or aromatic forms. Urea based synthetic CKs such as 6-(2-chloro-4-pyridyl)-N′-phenylurea (CPPU) and thidiazuron (TDZ) are often recognized as a third group of CKs (Schmülling 2004). For a long time, the isoprenoid CKs have received more detailed attention than the aromatic ones which is probably due to their relative higher abundance in higher plants, better knowledge of their biosynthesis and metabolism and the notion that aromatic CKs such as benzyladenine (BA) were synthetic (Van Staden and Crouch 1996). In terms of their functions, Holub et al. (1998) postulated that the isoprenoid CKs have a more potent effect on growth processes involving continuation of the cell cycle while aromatic CKs exert a greater influence on developmental processes, especially those involving morphogenesis and senescence. Cytokinins are usually present in low concentrations (micro molar) and exist endogenously in the form of free bases, nucleosides, glucosides and nucleotides in plants (Schmülling 2004; Doležal et al. 2007). In other words, CKs form reversible conjugates with sugars and amino acids (Bajguz and Piotrowska 2009). These conjugates act as storage, transport or biologically inert forms of CKs which are responsible for the physiological and developmental plasticity observed in plants (Letham and Palni 1983; Bajguz and Piotrowska 2009). Although the adenine-based forms of CKs are mostly preferred in PTC, the synthetic CKs such as TDZ and CPPU are also useful, especially in developing protocols for recalcitrant genotypes (Murthy et al. 1998). The effectiveness of these synthetic CKs has been demonstrated in micropropagation of apple (Fasolo et al. 1989), banana (Makara et al. 2010), sugar cane (Vinayak et al. 2009) and orchids (Mata-Rosas et al. 2010).

Despite the presence of numerous natural and synthetic CKs for use in PTC, the search for new CKs has remained active although not many laboratories are committed to it. The major advances in chromatographic and spectrometric techniques has aided the search (Tarkowski et al. 2009). Most importantly, the ever growing application of plant hormones in plant growth and improvement as well as the physiological and economic (to a lesser extent) limitation of the existing CKs have spurred the active search for new CKs (Kamínek 1992; Strnad 1997; Werbrouck 2010). Consequently, a new class of aromatic CKs, the topolins were discovered and identified. In the topolin basic structure, the presence of an hydroxyl group on the benzyl ring differentiates it from BA. Although the first isolation and identification of hydroxylated BA (oT) from Populus robusta dates back to the early 1970s (Horgan et al. 1973; Horgan et al. 1975), Strnad et al. (1997) are credited with the isolation of mT [6-(3-hydroxybenzylamino)purine], the highly active hydroxylated BA derivative from mature leaves of Populus × canadensis. Thereafter, four new methoxy derivatives of the topolins identified as 6-(2-methoxybenzylamino)purine (ortho-methoxytopolin, MeoT), 6-(3-methoxybenzylamino) purine (meta-methoxytopolin, MemT) and their 9-β-d-ribofuranosyl derivatives (MeoTR and MemTR) were isolated from Arabidopsis thaliana and Populus × canadensis (Tarkowská et al. 2003). A recent review has subsequently listed different topolins from a few other plant species (Subbaraj 2011). These major breakthroughs coupled with favourable results from the activity evaluation experiments using various bioassays and PTC systems have stimulated the increased use of these new CKs.

Countless PTC protocols have been developed and optimized for many plant species. These protocols are developed and optimized based on quantitative and qualitative evaluation of growth parameters such as shoot multiplication rate, rooting, plant appearance and acclimatization competence. The choice of medium type such as Murashige and Skoog (MS), gelling agent as well as type and concentration of PGRs are based on these essential growth parameters. Growth, by and large, is a function of cell division and it is the factors controlling this process that receive major emphasis when developing a PTC protocol. With the exception of a few laboratories worldwide, PTC experiments aimed at an in-depth understanding of the physiological and biochemical mechanisms by which PGRs bring about the changes are considered only when things do not work and/or when tissue abnormalities are encountered. As a result, many scientific reports on PTC lack basic metabolic and physiological explanation of growth. Recently, we proposed an improved or alternative approach whereby the type and concentrations of PGRs (and some other components) used in the cultivation media are optimized not only based on essential growth and acclimatization parameters but also correlated with optimum levels of active endogenous PGRs as well as minimization of inactive toxic metabolites (Malá et al. 2009; Bairu et al. 2011b). We believe that plant tissue culturists need to have a basic understanding of plant hormone physiology in relation to PTC. Hence, brief overviews on the CK biosynthesis, metabolism and structure–activity relationships are discussed.

CK biosynthesis and metabolism in relation to plant tissue culture

Plant tissue culture entails the growth of plants in artificial environments involving the use of externally applied PGRs such as CKs and auxins. This process interferes with, or influences the natural process of CK biosynthesis, a process little understood despite considerable attention and work on plant hormone physiology (Hutchison and Kieber 2002; Pan et al. 2008; Jiang and Guo 2010). Taylor et al. (2003) identified factors such as the extremely low levels of endogenous CKs, the central role of the likely precursors in cellular metabolism, the existence of numerous native substances with more or less pronounced CK-like activity and the reliance on incorrect or only partially correct assumptions about CK biosynthesis, as obstacles to the understanding of the biosynthetic pathways.

Recent findings on CK analysis of various in vitro grown species such as Watsonia lepida (Ascough et al. 2009) and Harpagophytum procumbens (Bairu et al. 2011b) indicated variation in type and composition of CK metabolites suggesting that plants vary in their CK metabolism patterns. According to Palmer et al. (1981) and Van Staden and Crouch (1996), the variation in the metabolism of applied CKs may be attributed to differences in the stages of development, physiological condition of the plant, organ type, plant species analyzed, concentration of applied PGRs and methods of application. Variation in the composition of CK metabolites of normal and necrotic shoots of in vitro grown Harpagophytum procumbens is a notable example of the effect of physiological condition on CK metabolism (Bairu et al. 2011b). This variation therefore makes it difficult to draw a general conclusion about the role of CK metabolites in plant growth and development in vitro. The variation in the composition of CK metabolites may also indicate that it is unlikely for all the CKs to be converted to a common metabolite responsible for the growth responses observed.

Plant tissues metabolize exogenous CKs to different types of metabolites such as products of ring substitution (ribosides, nucleotides, N-glucosides) and products of side chain substitution (O-glucosides) or cleavage (adenine, adenosine, adenosine-5′-monophosphate) (Letham and Palni 1983; Van Staden and Crouch 1996). Although the functional significance of these metabolites is still not well understood (Wagner and Beck 1993; Van Staden and Crouch 1996), long standing suggestions by Letham and Palni (1983) indicate that these compounds could be: active forms of CK, i.e. the molecular form that induces growth or physiological response; translocation forms; storage forms which would release free (active) CKs when required; detoxification products formed due to the application of exogenous CK at toxic levels; deactivation products formed to lower endogenous (active CK) levels; and post-activation products, the formation of which is coupled with CK action.

The structural variation among CKs is another aspect contributing to the difference in CK metabolism. Van Staden and Crouch (1996) emphasized the absence of a common metabolic pattern for aromatic CKs compared to the knowledge on isoprenoid CK metabolism. The need for better understanding of the pattern of CK metabolism triggered the study of CK action at molecular and cellular levels with emphasis on signal perception, transduction and response (Kamada-Nobusada and Sakakibara 2009). Initially, these studies were mainly done by searching for proteins which may serve as signal receiving molecules followed by studying the signal transduction pattern(s) (Strnad 1997). Although very little is known about the molecular mechanism by which target cells for PGRs translate the signals to specific responses, advances in molecular genetics are making such studies at protein and gene level possible (Kamada-Nobusada and Sakakibara 2009). Earlier studies by Libbenga and Mennes (1995) indicated that in a hormonal system, cells of different tissues and organs not only transmit signals, but are also capable of detecting signals which they receive from other parts and respond to those signals in their own characteristic way.

Strnad (1997) discussed the specificity and complexity of CK binding. The existence of two related groups of CK binding proteins (CBPs), one with low affinity to Z but with strong affinity to BA and the other with an opposite character, serves as an indication that there is a distinction between aromatic and isoprenoid CKs with respect to nature of binding and receptor response. In line with the proposal by Iwamura et al. (1980), the differences found in the electron structure and hydrophobicity of the N6-substituent are probably related to the CK binding site; the role of modified adenines in CK-binding interaction has been elucidated using X-ray crystallographic structural studies (Strnad 1997).

Mok et al. (2005) reported that topolins and hydroxylated TDZ derivatives are substrates for CK O-glucosyltransferase with position specificity related to receptor recognition. It was shown that mT and oT derivatives are preferred substrates of zeatin O-glucosyltransferase (ZOG1; enzyme encoded by Phaseolus lunatus) and cis-ZOG1 (enzyme encoded by Zea mays), respectively. The authors reported a correlation between the activity of the CKs and their ability to serve as a substrate for glucosyltransferase, thereby suggesting a similarity between CK-binding sites on the enzyme and CK receptors. In addition, they found support for their interpretation from CK recognition studies involving the Arabidopsis CRE1/WOL/AHK4 and Zea mays ZmHK1 receptors. The AHK4 receptor responded to trans-Z and mT while the ZmHK1 receptors responded to cis-Z and oT.

Cytokinins, in combination with auxins, are known to affect the basic mechanisms of cell proliferation and differentiation (Faure et al. 1998). Harrar et al. (2003) demonstrated that this hormonal control of cell proliferation and differentiation requires PASTICCINO (PAS) genes. The authors determined the role of the PAS gene by analysing the expression profiles of several genes involved in cell division and meristem functioning and found that differentiated and meristematic cells of the PAS mutants were more competent for ectopic cell division, and were especially enhanced by CKs. They further demonstrated that disorganised cell divisions were associated with the deregulation of cell cycle marker genes like cyclin-dependent kinase A (CDKA) and cyclin B1 (CYCB1).

These variations in CK metabolism, binding and signal perception could possibly explain the inherent variation in response among plants in PTC systems. In addition, the structural variation among CKs is undoubtedly another avenue worth investigating because the literature indicates that a slight variation in structure results in a major change in activity in tissue culture systems. Hence, factors affecting CK structure–activity relationships are discussed below.

CK structure–activity relationships

The concept of CK structure–activity relationship has attracted the attention of researchers for more than five decades. Today, research interest and focus on the same topic is as topical as previously, due to the growing application of our knowledge of PGRs. It is well documented that the activity of CKs is highly affected by their structure; with the structure activity relationship of CKs being affected in a number of different ways (Haberer and Kieber 2002). The influence of structural modifications on the activity of natural and synthetic CKs has been extensively studied. Different factors such as using: an intact adenine moiety with an N6-substituent of moderate molecular length (Skoog and Armstrong 1970), an intact purine ring, unsubstituted 1- and 3-positions, optimum side chain length of five carbon atoms, unsaturation in the side-chain, a substituent which is planar and hydrophobic (Hecht et al. 1970; Nishikawa et al. 1986), an electron rich nitrogen group located opposite to the substituent and linkage atoms or a group that connects the purine ring with the side-chain and restrict the molecular configuration (Nishikawa et al. 1986), 4-hydroxylation of the iso-pentenyl side chain (Matsubara 1980) have been mentioned as a structural requirement for high CK activity. Alternatively, Chen and Kristopeit (1981) suggested that activity depends on the interconversion between the free bases and their ribosides. Matsubara (1980) attributed the structure activity relationship of CKs to several factors which could be placed in one or more of the above mentioned categories. General structural and functional requirements for CK activity are discussed briefly.

The general effect of ring substitution of aminopurines

Most of the pioneering research on the effect of ring substitution on CK activity dates back to the 1950s, 1960s and 1970s. A detailed and comprehensive review of this effect was done by Matsubara (1980). Replacing the furfuryl group of KIN by a wide range of other side-chains was possible without any considerable loss of CK activity. BA, arguably the most utilized CK in PTC owes its discovery to this form of ring substitution—replacing the furfuryl group by a benzyl group. BA was found to be more active than KIN in the tobacco callus assay. The activity of ring substituted aminopurines is also affected by the degree of saturation. Generally, aminopurines with an unsaturated ring are more active than the saturated ones (Skoog et al. 1967; Skoog and Armstrong 1970; Matsubara 1980).

The effect of the hydroxyl group on the side chain

Hydroxylation of the trans methyl group in the N6 side-chain of N6-(Δ2-iso-pentenyl) adenosine increased its biological activity but the activity was not affected or decreased when the cis methyl group was hydroxylated. This targeted hydroxylation of aromatic CKs is believed to regulate their activity (Kamínek et al. 1987). This form of hydroxylation is likely to have an effect on the dihedral angle and the formation of an hydrogen bond with the nitrogen at N1 position—both known to affect the activity of aromatic CKs depending on the position of the hydroxyl group (Trávníček et al. 1997; Nirmalram et al. 2011).

The position of the hydroxyl group on the side-chain has a significant effect on the biological activity of the parent CKs. Substitution on the phenyl ring enhanced the activity in the order of meta > ortho > para in most bioassays for purinyl and urea-type CKs (Kamínek et al. 1987). Horgan et al. (1975) noted a decrease in activity of N6-benzyladenosine after hydroxylation of the phenyl ring at the ortho position. Mok and Mok (1985) also demonstrated that activity was in order of meta > ortho > para in the Phaseolus lunatus callus assay for the hydroxylated derivatives of TDZ. The positive reports on the role of mT in PTC (see Table 1 for references) could therefore be attributed to these structural changes.

The effect of a double bond on the side chain

Hecht et al. (1970) highlighted that the presence of a double bond on the side chain is one of the structural requirements for high CK activity. A high CK activity can also be achieved or maintained by keeping side-chain planarity. Furthermore, the authors postulated that the addition of a substituent to the double bond disturbed the side-chain planarity, thereby reducing CK activity. The position of the double bond also affects CK activity. Leonard et al. (1968) demonstrated that the 2,3-position of the side-chain as in 6-(3-methyl-2-butenylamino)purine enhanced activity while shifting the double bond to the 3,4-position as in 6-(3-methyl-3-butenylamino)purine reduced the activity in the tobacco callus bioassay. By synthesizing and testing the biological activity of two 6-substituted purines, one with an ά-double bond and one without, Nishikawa et al. (1986) revealed that the compound with the ά-double bond was twice as active as the one without, indicating the importance of the double bond in CK activity.

The effect of the methyl and amino groups on the side chain

Cytokinin activity is affected by the methyl group(s) and their position on the side-chain of N6-substituted adenines. Lengthening the bridge or removal of the methyl group between the ring and the amino group on the 6-position of the purine ring in ring-substituted adenines decreases CK activity (Matsubara 1980). Alternatively (at least in the case of aromatic CK), high ability of these anilinopurines to inhibit CK oxidase-dehydrogenase, the key enzyme of CK catabolism, and thereby to increase levels of endogenous active CKs in the treated plant tissues has been recently shown (Zatloukal et al. 2008). Di-substitution with two methyl groups at position 3- produced a highly active 6-(γ,γ-dimethylallylamino)purine whereas addition of a methyl group to the 1, 2, or 3-positions of the carbon atom on the side-chain did not affect the activity. However, putting these two methyl groups at position 1- of the side-chain reduced the activity a 100 fold. Shifting the methyl group from position 3- to 2- of the side-chain of cis and trans-Z or removing it from cis-Z resulted in a significant decrease in activity (Skoog et al. 1967; Matsubara 1980). The replacement of the amino group of KIN by a sulphur atom resulted in a considerable decrease in activity suggesting the importance of the amino group on the N6-position for CK activity (Matsubara 1980).

The effect of side chain configuration (geometrical and optical isomers)

Studies have shown that the side-chain configuration affects CK activity. For example, trans-Z was more active than the cis isomer in stimulation of cotyledon expansion, retention of chlorophyll in detached leaf pieces, induction and stimulation of chlorophyll synthesis in Cucumis sativus and betacyanin synthesis in Amaranthus caudatus seedlings grown in the dark (Kamínek et al. 1979). The difference in the side-chain configuration of geometrical isomers affects the side-chain planarity, which in turn affects activity (Hecht et al. 1970). The interaction between the CK molecule and its receptor site is also influenced by the absolute configuration around the asymmetric carbon (Matsubara 1980).

The effect of position of the substituent on the purine ring

The position of the substituents in the purine ring affects CK activity and gives the CKs their respective identities. For a CK to remain active, the reactive N1 position must remain free (Matsubara 1980). This concept elaborates the fact that oT is less active than mT due to the formation of the hydrogen bond with the nitrogen at N1 position (Holub et al. 1998).

The effect of substitution at N2 varied depending on the type of assay and the nature of the substituents. An example of a compound with substitution at N3 position is triacanthine—a naturally occurring adenine derivative. Similar to the isomers with N1 substitution, triacanthine was active only following autoclaving leading to the postulation of producing N6-substituted adenine (Leonard and Henderson 1975). These authors demonstrated that when autoclaved, 3-substituted adenines undergo rearrangement and conversion in low quantity to N6-substituted adenines. The N6 substituted adenines are the most active of all N-substituted adenines and generally, the mono-substituted ones are more active than the corresponding di-substituted ones (Skoog et al. 1967). In addition, substitutions at positions N7, N8 and N9 produce inactive or slightly active products depending on the nature of the substituent molecules (Matsubara 1980).

The effect of an halogen substituent

The activity enhancing effect of halogen substitution on isoprenoid (Clemenceau et al. 1996; Haidoune et al. 1998) and aromatic (Bogaert et al. 2006; Doležal et al. 2006; Doležal et al. 2007) CKs are well documented. Different authors explained the activity enhancing effect of halogen substitution differently, for example: the high electronegativity and small size of the halogens (Crocker et al. 2007), the ability of organo-halogen compounds to form structural motifs via inter-molecular interactions (Emmerling et al. 2007), decreased sensitivity of the compounds to CK oxidase (Clemenceau et al. 1996; Haidoune et al. 1998), and ability to form an hydrogen bond with electron donors of a CK receptor (Doležal et al. 2007) are proposed to be the contributing factors.

The structural variability of organic compounds in general and CKs in particular offered a wide range of possibilities to modify their structure and enhance their application spectrum. The variations in the biological activity of the same CKs in various bioassay systems are logical indications of the diverse recognition patterns and/or signalling mechanisms that may operate in CK-dependent physiological responses. It also highlights the possibility of designing specific compounds to modulate a particular CK-dependent response (Doležal et al. 2006; Doležal et al. 2007). This makes interpretation of biological activity test results to be dependent on the nature of the assay and plant species used.

While the above structure–activity relationships and considerations provide an insight on the factors that affect CK activity, these also probably help us understand and/or explain some of the growth related observations in PTC systems. Therefore, we examine the role of topolins in PTC systems in comparison to other CKs.

Effect of topolins on shoot multiplication and elongation

Shoot regeneration and multiplication during micropropagation is affected by the type and concentration of applied PGRs, especially the CKs due to their importance in cell division and organogenesis (Howell et al. 2003). Benzyladenine remains one of the most effective CK that promotes in vitro shoot regeneration and multiplication of plant species. In an attempt to address various physiological and developmental problems associated with CKs in general and BA in particular however, other CKs are being tested.

As shown in Table 1, various studies on a wide range of plant species highlighted the potential of topolins as a substitute to the commonly used CKs. Using equimolar concentration (10.0 μM) of either mT or BA, Werbrouck et al. (1996) observed better shoot–root balance with mT-treated Spathiphyllum floribundum plantlets. Similarly, mT at various concentrations produced more shoots compared to either BA or Z during the micropropagation of Aloe polyphylla (Bairu et al. 2007). In banana cv. ‘Williams’, the use of topolins (mT, mTR, MemT, MemTR) at 7.5, 15 and 30 μM had higher shoot multiplication rates than BA (Bairu et al. 2008). Studying the effect of nine CKs on shoots production in micropropagated Curcuma longa, Salvi et al. (2002) obtained more shoots with the use of kinetin ribosides (KINR), iP or mT (no significant difference) compared to either BA or other CKs used. However, regenerated shoots from mT treatment were greener and stouter than KINR and iP-treated one.

Nevertheless, unfavourable responses from the use of topolins for shoot regeneration and multiplication rate have also been observed in certain plant species. In Rosa hybrida for example, Bogaert et al. (2006) reported that MemTR-treated explants had a lower multiplication rate than the BA treatment. A number of similar responses (where other CKs were more effective) have been documented by several authors (Table 1). There was no definite pattern in the plant species responses to the different CKs; an indication of the interaction of several factors during shoot regeneration and multiplication phase. Factors such as genotype, CK concentration used and medium type (Dobránszki et al. 2004; Vinayak et al. 2009; Magyar-Tábori et al. 2010; Wojtania 2010) remain vital and need to be taken into consideration while optimizing PTC protocols.

Effect on rooting and ex vitro acclimatization

Lack or inadequate rooting is among the major problems in PTC. In micropropagated plants generally, there is a positive correlation between good rooting (mainly stimulated by auxins) and their ability to acclimatize faster (Werbrouck et al. 1996; Koetle et al. 2010). Acclimatization of in vitro grown plants to natural conditions is a vital step in the micropropagation of many species especially for large scale application of in vitro techniques As opposed to the ex vitro conditions, during in vitro micropropagation, plants grow under constant temperature, high relative humidity, low irradiance, sugars as carbon source, PGRs in nutrient medium, variable and often insufficient CO2 concentrations (Pospíšilová et al. 2007). These differences result in variation in factors such as leaf structure, water relations and photosynthetic parameters which are critical for the successful acclimatization of in vitro plants (Pospíšilová et al. 1999). In addition, the effect of the applied PGRs such as CKs, auxins and abscisic acid are crucial in the successful and improved acclimatization of in vitro plants. Several studies on different species have reported that the type and concentration of CKs have a profound effect on in vitro plant acclimatization competence (Moncaleán et al. 2001; Bairu et al. 2008; Valero-Aracama et al. 2010). In addition, some authors have reported that CKs generally have inhibitory effects on rooting, resulting in poor acclimatization rates afterwards (Werbrouck et al. 1995; Bairu et al. 2008).

Although application of exogenous CKs enhances shoot formation during micropropagation, the CK BA is often associated with negative side-effects during acclimatization of some plant species. According to Werbrouck et al. (1995), these harmful effects are partly due to the formation of N-glucosides or alanine conjugation, which are biologically inactive (characterize by slow release of active CK free bases) and chemically stable metabolites. More importantly, BA-treated plantlets have the tendency of accumulating these toxic BA metabolites in their basal (rooting zone) portions. Consequently, these metabolites interfere with rooting and acclimatization competence in micropropagated plantlets (Werbrouck et al. 1995). As demonstrated by Valero-Aracama et al. (2010), BA treatment caused detrimental biochemical, physiological and developmental effects in Uniola paniculata cultures that resulted into reduced acclimatization competence. The authors discovered that similar problems were absent in mT treatments; which is a further indication of the relative lower toxicity of mT over BA and its analogues. Concomitantly, Werbrouck et al. (1996) postulated that mT metabolites are less stable and produce reversibly sequestrated metabolites. The presence of an hydroxyl group in topolins give them a structural advantage over BA in that they can undergo O-glucosylation to form storage forms (Bairu et al. 2011b).

In contrast to BA, mT stimulated in vitro rooting activity in Spathiphyllum floribundum (Werbrouck et al. 1996). Similarly, Bairu et al. (2007) observed that both mT and mTR promoted the rooting of Aloe polyphylla shoots in the multiplication medium and approximately 90% of plantlets treated with mT acclimatized successfully compared to a 65% survival rate recorded with BA-treated plantlets. Low concentrations (0.005–0.010 mg l−1) of mTR significantly increased the rooting and survival of Solanum tuberosum cv. Jaerla plantlets (Baroja-Fernández et al. 2002). However, the authors observed a decrease in the survival rate as the concentration of the mTR was increased.

The positive effect of topolins on rooting, however, is not universal. Meta-topolin and mTR at 22.2 μM had an inhibitory effect on rooting of Musa spp. (cv. Williams) compared to BA (Bairu et al. 2008). Valero-Aracama et al. (2010) observed that mT at 10 μM or higher had inhibitory effects on the in vitro rooting of two Uniola paniculata genotypes. Similarly, Escalona et al. (2003) in a study on plantain cv. ‘CEMSA 3/4’ reported a progressive reduction in root production as the concentrations of BA and mT were increased (1.33–22.2 μM), however, 1.33 μM mT gave the highest number of roots. The general trend indicates that higher concentrations of most topolins (as is the case with most CKs) were detrimental to the rooting and acclimatization competence of micropropagated plants. Hence, a wide range of concentrations should be incorporated while investigating the use of these topolins. The presence of well-developed roots in multiplication media and inhibition of rooting in some species warrant detailed investigation on the effect of topolins in the physiology of root development.

The interaction of CKs with auxins has always been an important consideration in PTC. Very little published work on topolins is available on this aspect. Recently, we observed the differential effect of the role of CK-auxin interaction on endogenous CK levels in relation to shoot-tip necrosis (STN) (Bairu et al. 2011b). It was noted that the presence of the auxin, indole-3-acetic acid (IAA) in the culture medium enhanced the formation of 9-glucosides in BA-treated cultures but reduced it in topolin-treated cultures. Better O-glucosylation was also observed in topolin-treated cultures when IAA was omitted. Similarly, Malá et al. (2009) measured lower levels of IAA in BA-treated explants of Sorbus torminalis compared to mT—an explanation to reduced rooting by BA-treated cultures. Although the structural differences will have a role to play, the physiological and biochemical events leading to these effects are yet to be fully understood.

Effect of topolins on countering in vitro abnormalities

Effect on hyperhydricity

Hyperhydricity is a morphological and physiological disorder observed in micropropagated plants (Debergh et al. 1992). It results from either a passive diffusion of water into the tissue or an active phenomenon related to strong metabolic disturbances (Pâques 1991). Hyperhydricity is a common problem that has been observed in micropropagated plants and hinders the potential of in vitro techniques for mass propagation of plant species (Kevers et al. 1984).

Plant growth regulators, the type and concentration of the gelling agents, presence of large quantities of mineral nutrient (mainly NH +4 and Cl−) ions in the medium and high relative humidity in the culture vessels are among the factors causing hyperhydricity (Pâques and Boxus 1987; Franck et al. 2004). Although the mechanism and interactions of these factors are still not well understood, the effect of PGRs in general and CKs in particular are well documented.

Bairu et al. (2007) investigated the effect of topolins on hyperhydricity in Aloe polyphylla. At an optimum concentration of 5.0 μM, there were no hyperhydric shoots in topolin (mT and MemT)-treated plants. Although all the CK treatments caused hyperhydricity at higher concentrations, it was most severe with BA treatments. Similarly, drastic alleviation of hyperhydricity was also reported with the use of a low concentration of mT for Beta vulgaris (Kubalákova and Strnad 1992) and for several Malus × domestica cultivars (Dobránszki et al. 2002; Dobránszki et al. 2004; Dobránszki et al. 2005).

Effect on shoot-tip necrosis (STN)

Shoot-tip necrosis (shoot die-back) is a common physiological disorder in micropropagated plants (Bairu et al. 2009c). The abnormality is often accompanied by browning of buds and leaves eventually leading to plant death. The symptoms result from the senescence and death of tissues in the apical bud which subsequently proceed basipetally (Kataeva et al. 1991). Shoot-tip necrosis is a consequence of a complex set of factors (Bairu et al. 2009c), the effect of PGRs especially CKs however, remains dominant (Il’ina et al. 2006). Although the rate of STN increased with increasing CK concentration, Bairu et al. (2009a) observed significantly lower STN in micropropagated Harpagophytum procumbens supplemented with mTR compared to BA treatment. Similarly, the use of MemTR alleviated STN associated with micropropagation of Barleria greenii (Amoo et al. 2011).

Effect on early senescence

Senescence is a genetically programmed process that involves a general termination of cellular structures, followed by mobilization of the degradation products to other parts of the plant (Woo et al. 2004). Leaf senescence is a critical process for the fitness of plants and plays a major role in evolution of plants (Nam 1997). Generally, it is controlled by many internal and external factors (Noodén 1988). Plant growth regulators such as CKs, abscisic acid, auxin and ethylene do play some roles in regulation of leaf senescence. It is understood that CKs are one of the main regulators of senescence in plants (Van Staden et al. 1988). Cytokinins are known to delay aging in many plant organs through protein breakdown inhibition as well as the stimulation of RNA and protein synthesis (Palavan-Ünsal et al. 2002b). Molecular studies have provided evidence on the effect of CKs on senescence (Gan and Amasino 1996). Yet, the molecular mechanisms of CK action as well as the molecular basis of plant senescence generally are poorly understood.

Early senescence is a common problem in PTC of many plant species, for example, Mithila et al. (2001) highlighted this problem as hindering the micropropagation of many Pelargonium cultivars. Subsequently, Wojtania (2010) discovered that mT was better than BA in the inhibition of early senescence in seven Pelargonium cultivars. Meta-topolin-treated Pelargonium plantlets had higher chlorophyll content. Likewise, the application of mT to leaf segments of Triticum aestivum retarded senescence by decreasing protease activity and chlorophyll loss (Palavan-Ünsal et al. 2002b). Further studies also established that both nitrogen and polyamine contents increased due to application of mT. As a result, the authors suggested that increased peroxidase activity, polyamines and nitrogen contents in mT-treated excised Triticum aestivum was a contributing factor to the overall anti-senescence activity of mT (Palavan-Ünsal et al. 2004). Using Rosa hybrida as a model plant, Bogaert et al. (2006) investigated the anti-senescence effect of BA and topolins on micropropagated plants. Although no CK treatment prevented the older (lower) leaves from senescing after 6 weeks, MemTR treatment showed the most noteworthy anti-senescing activity. Furthermore, after 18 weeks, 50% of MemTR-treated plants were alive and the next active CK, 6-(3-fluorobenzylamino)purine-9-riboside (FmTR) had only a 14% survival rate.

Foliar application of mTR to field grown sugar beet delayed senescence with increased content of natural cytokinins as well as higher yield compared to the untreated control (Čatský et al. 1996). These effects were associated with the reduced respiration and/or stimulated membrane transport processes in the mTR-treated plants (Kotyk et al. 1996). In view of the importance of early senescence inhibition, molecular approaches have been suggested for better understanding of the process (Gan and Amasino 1997). In addition, we recommend both in vitro and ex vitro application of these topolins, to evaluate their effect on retarding senescence on various plant species.

Effect on histogenic stability

Histogenic stability is an important factor in commercial micropropagation of plants for two contrasting purposes, namely preserving pre-existing variation (chimeras) and preventing unnecessary (somaclonal) variation, see Bairu et al. (2011a) for a detailed review. In addition, incidence of somaclonal variation serve as a potential avenue for generating new variants in ornamental potted plants such as Anthurium (Winarto et al. 2011).

During micropropagation of ornamentals, it is essential to maintain the pre-existing genetic composition of chimeric plants for its commercial value. Presently, KIN is the only tool to slowly and safely regenerate these valuable ornamentals (Bogaert et al. 2006). However, KIN being a weak CK is often associated with lower shoot multiplication (Amoo et al. 2011). These drawbacks probably stimulated the application of topolins in the micropropagation of chimeras. Maintaining the histogenic stability of Petunia meristem was improved using MemTR when compared to BA. Furthermore, application of MemTR resulted in a superior visual quality of the axillary shoots (Bogaert et al. 2006).

An increase in BA concentration increased the rate of variation in Cavendish banana cv. ‘Zelig’ (Bairu et al. 2006). In another study, Bairu et al. (2008) reported the absence of significant differences in the occurrence of somaclonal variation when BA, mT and mTR were compared for ‘Williams’ banana. The authors postulated that the results obtained in their investigation could have been influenced by carry-over effects of BA from the initial cultures due to its effect on somaclonal variation (Bairu et al. 2006). A similar scenario was observed during the micropropagation of Barleria greenii that was initially maintained on BA-supplemented media (Amoo et al. 2011). The authors observed that the abnormality index of mTR and MemTR treatments were lower than the control treatment. Thus, it was suggested that the observed abnormality index in mTR and MemTR treatments was a carry-over effect from BA. Presently, stringent studies to investigate the role of topolins on somaclonal variation by minimizing the influence of BA as a confounding factor on the genetic stability of ‘Williams’ bananas are on-going in our laboratories.

Concluding remarks and future considerations

The choice of CK is amongst the most critical factors in developing a successful PTC protocol. Numerous examples from Table 1 have shown that topolins can successfully replace the commonly used CKs in many PTC protocols. They can also play corrective roles on some physiological disorders. It should however, be noted that there are species that respond better to CKs other than the topolins; hence topolins should not be taken as a panacea and must pass through the routine process of selection. While it is essential to optimize efficient PTC protocols through stringent choice of CKs, tissue culturists should also give emphasis to the associated physiological and metabolic events taking place in culture during the optimization process so that they contribute towards better understanding of the mode of action of these molecules. Such an approach will help us solve associated physiological and developmental problems in vitro. Since many tissue culture laboratories are not well equipped for physiological experiments, collaborative efforts could help bridge the gap.

We recommend that fresh cultures are started to investigate the role of topolins to avoid possible carry-over effects of other CKs and culture additives thereby eliminating reaching incorrect conclusions. To elucidate the broad spectrum of action of topolins, our research group at the Research Centre for Plant Growth and Development, University of KwaZulu-Natal in South Africa in collaboration with the Laboratory of Growth Regulators in the Czech Republic are testing these CKs on a wide range of species.