Euphytica 37: 31-36 (1988)
The induction in vitro of adventitious shoots in Rosa
Davina Lloyd1, Andrew V. Roberts1 and Keith C. Short2
1 Department of Biology and Biochemistry, North East London Polytechnic, London E15 4LZ, England;
2 Department of Life Sciences, Trent Polytechnic, Clifton Lane, Nottingham, NG11 8NS, England

Received 5 January 1987; accepted in revised form 16 March 1987

[For American readers, mg l-1 = mg/liter]


Adventitious shoots were formed on excised leaves, roots and callus of Rosa persica X xanthina and on excised leaves of R. laevigata and R. wichuraiana on culture media that included BAP and NAA as growth regulators. Shoots formed freely on freshly cultured callus of R. persica X xanthina but their production declined in successive cultures and ceased after twelve weeks. Transplantation to soil was improved by rooting plantlets in cellulose plugs in vitro and transferring plantlets to soil while still in the plugs.


There are numerous reports of the culture in vitro of embryos, cells, callus and shoots of Rosa (Skirvin & Chu, 1979). The multiplication of shoots has been reported on media containing salts and vitamins (Murashige & Skoog, 1962) supplemented with 30-50 g l-1 sucrose, 6-8g l-1 agar, 0.5-0.3 mg l-1  benzylaminopurine (BAP) and either 0.004-0.3 mg l-1 naphthaleneacetic acid (NAA) or 0.05-2.0 mg 1-1 indoleacetic acid (IAA) (Skirvin & Chu, 1979; Hasegawa, 1979, 1980; Bressan et al., 1980, 1982; Davies, 1980; Hyndman et al., 1982; Khosh-Khui & Sink, 1982). Plants which were propagated in this way showed morphological uniformity after transplantation to soil, and were probably all derived from primary shoot meristems.

The only investigation which indicated that shoots of adventitious origin might have been generated is that of Hill (1967). He took stem explants from a Hybrid Tea rose and induced callus on an unspecified synthetic medium which included either 0.2 mg l-1 kinetin and 2.0 mg l-1 NAA, or 2.0 mg l-1 kinetin and 0.2mg l-1 NAA. Shoots were formed when this callus was transferred to a medium which included 1mg l kinetin. It seems possible that the shoots which Hill (1967) observed were not necessarily of adventitious origin, as they might have arisen from primary meristems that had retained their identity despite the unorganised proliferation of adjacent cells. In the present investigation, care was taken to ensure that the induction of adventitious shoots was tested on explants that did not include primary shoot meristems. Excised leaves, roots and internodal segments were accordingly used.

Rosa persica (Michx. ex Juss.) Bornm. (= Hulthemia persica Michx. ex Juss.) is adapted to a desert habit and has simple leaves and a tap root. Its flowers have a unique 'red eye' on a yellow background which is of great interest to rose breeders. Like all other hybrids between R. persica and other roses, the hybrid R. persica X R. xanthina Lindl. is sterile (Roberts, unpublished). It is predictable that if explants of R. persica X xanthina were to be treated with colchicine and if adventitious shoots were subsequently induced, a proportion of these shoots would be pure (non-chimeral) amphidiploids. It is possible that these amphidiploids would be fertile. With this prospect in mind, the primary objective of the present investigation was to assess the ability of R. persica X xanthina to generate adventitious shoots. R. hybrida L. 'Clarissa', R. hybrida L. 'Dame of Sark', R. rugosa L. Thunb. 'Scabrosa', R. wichuraiana Crép. and R. laevigata Michx. were included in this study to meet a secondary objective, which was to provide a more broadly based assessment of the ability of roses to generate adventitious shoots.

Materials and methods

Plant materials

Specimens of R. persica X xanthina, R. hybrida 'Clarissa', R. hybrida 'Dame of Sark' and R. rugosa 'Scabrosa' were donated by Harkness New Roses Ltd. Specimens of R. wichuraiana and R. laevigata were obtained from Hillier Nurseries, Romsey, England.

Culture in vitro

Terminal buds were excised from vigorously growing shoots and immersed for 20 min in a solution of sodium hypochlorite containing 1.0-1.5 per cent available chlorine and 0.1 per cent 7X surfactant (Flow Laboratories Ltd.). Buds were then immersed in two changes of sterile distilled water, each for 20 min and inoculated onto multiplication medium. Multiplication medium consisted of MS salts and vitamins (Murashige & Skoog, 1962), 40g l-1 sucrose, 0.5 mg l-1 benzylaminopurine (BAP) and 0.005 mg l-1 naphthaleneacetic acid (NAA). This was adjusted to pH 5.6, 8g l-1 agar was added and the medium was autoclaved at 121°C for 15 min. The rooting medium which was finally adopted consisted of MS salts and vitamins diluted to 1/4 strength, 40g l-1 sucrose and 0.1 mg l-1 IAA. This was adjusted to pH 5.6, filter sterilized and added to radiation sterilized Sorbarod plugs (Baumgartner Papiers SA, Lausanne, Switzerland) which consisted of cylinders (length: 15 mm, diameter: 12 mm) made of laminated cellulose in a sleeve of perforated paper. Sufficient medium was added to saturate the plugs but in order to maximize aeration, no surplus was left standing in the culture vessel. Multiplication and rooting of plantlets in vitro was carried out in screw-cap vessels of 60 ml capacity. Induction of adventitious shoots was tested on media similar to the multiplication medium except that six levels of BAP, ranging from 0-5mg l-1, and six levels of NAA, ranging from 0-0.3mg l-1 were tested in all combinations. The internodal segments, roots and leaves which were tested were taken from plants cultured in vitro. A minimum of 20 explants were tested on culture medium in petri dishes of 90 mm diameter. The petri dishes were sealed with Nescofilm (Nippon Shoji Kaisha, Ltd., Japan).

All cultures were maintained at 23°C in a 16h photoperiod provided by Grolux fluorescent lights giving 1.8 Wm-2. [Watts/sq. meter]

Light microscopy

Tissues were fixed and embedded in wax following the procedure of Sass (1958). Sections were cut at 7-8 m thickness and stained in 0.05 per cent aqueous solution of toluidene blue. Starch grains were identified by their birefringence in plane-polarized light.


Callus cultures

Production of adventitious shoots occurred on callus of R. persica X xanthina only. Shortly after the introduction of this hybrid into culture, callus could be induced on internodal segments on media containing 1.0-2.0 mg l-1 BAP and 0.1-0.3 mg l-1 NAA. Six months after its introduction into culture, equivalent callus production could be initiated on medium containing lower levels of growth regulators (0.5 mg l-1 BAP and 0.005 mg l-1 NAA). A steady rate of growth was maintained by dividing pieces of approximately 10mm diameter (200-300 mg fresh weight) into pieces of approximately 5 mm diameter at intervals of 4 weeks. This callus was friable and pale green to yellow in colour. When callus derived from recently initiated shoot cultures was transferred to medium containing 3.0 mg l-1 BAP and 0.1-0.3 mg l-1 NAA a harder and darker callus was formed, which gave rise to adventitious shoots after 4 weeks (Table 1 and Fig. 1A). When callus was obtained from shoot cultures that had been established at least six months previously, adventitious shoots could be obtained on media containing lower levels of growth regulators (1.0 mg l-1 BAP and 0.005 mg l-1 NAA). Sections showed that the shoots arose from meristemoids at or below the surface of the callus (Fig. 1C). Callus that was maintained for more than 3 passages from its initiation lost its organogenic potential. The cells contained starch grains that were not evident in organogenic callus. (Fig. 1B). Shoots were detached from the callus and transferred to multiplication medium. Those with trifoliate leaves and slender stems survived, while those with unifoliate leaves and thick fleshy stems leaked phenolic substances into the culture medium and could not be maintained. Considerable variation in habit and leaf shape (Fig. 2) was found amongst plants derived from adventitious shoots. This suggested the possibility of somaclonal variation, but the stability of the observed changes has not yet been rigorously investigated.

Callus induced from internodal segments of R. laevigata, R. wichuraiana, 'Clarissa', 'Dame of Sark' and 'Scabrosa' failed to produce shoots on any of the test media. Cells of these calluses, like non-morphogenic callus of R. persica X xanthina, contained numerous starch grains.

Leaf cultures

Shoots with upto 3 simple leaves were formed on excised leaves of R. persica X xanthina, R. laevigata and R. wichuraiana only (Table 1). They appeared within six weeks of inoculation onto medium containing 0.5 mg l-1 BAP and no NAA (Table 1). They were formed directly on the petiole and the midribs of leaflets without intervening callus. On transfer to multiplication medium, only shoots of R. persica X xanthina survived.

Root cultures

Roots were excised, cut into lengths of 20 mm and transferred to the test media. Adventitious shoots were formed on root segments of R. persica X xanthina, only (Table 1). They were induced on medium containing 2.0 mg l-1 BAP. Light microscopy of sections showed that they arose from the region of the pericycle, without intervening callus. These shoots were detached and transferred to multiplication medium.

Table 1. Numbers of adventitious shoots formed in one hybrid and two species of rose on three types of explant.
  Number of adventitious shoots per explant* (± SE), derived from:
  Callus Leaf Root
R. persica X xanthina    5.0 ± 2.67    2.6 ± 0.60    3.2 ± 0.45
R. laevigata 0 1.75 ± 0.33 0
R. wichuraiana 0    1.2 ± 0.68 0
* Based on a minimum of 20 explants on best medium.

Fig. 1. R. persica X xanthina cultured in vitro.

  1. Adventitious shoots arising from callus (x 10).
  2. Sections through non-morphogenic callus (x 200). Starch grains can be recognised by their birefringence in plane-polarised light.
  3. Section through morphogenic callus, showing an adventitious bud developed below surface (x 110).
  4. Shoot growing in cellulose plug before transplantation to compost (X 1.4).

Transfer to soil

Transfer of shoots of R. persica X xanthina from in vitro culture to compost proved to be difficult whether they arose from primary meristems or adventitious shoots. On rooting medium a single root was usually formed which was long and unbranched, like the tap root of R. persica. This was easily broken on transfer to compost and was not replaced. When 50 shoots were transferred from rooting medium solidified with 8 g l-1 agar, to Levington's compost in a mist propagator, none survived. However, when 20 shoots were rooted in vitro in Sorbarod plugs moistened with rooting medium without agar (Fig. 1D), 85 per cent survived transplantation to compost. This reduced mortality is attributed to the protection given to the roots by the plug and the availability, to the plant, of water held in the cellulose matrix of the plugs.

Fig. 2. Leaves from selected in vitro plantlets of Rosa persica X xanthina representing a range of morphological types (Bar line: 10mm).


A decline in the capacity of callus to regenerate shoots after repeated subculture is known in many genera (George & Sherrington, 1984) and was observed in R. persica X xanthina in this study. Starch grains accumulated in the callus of R. persica X xanthina during this decline and were conspicuous in non-organogenic callus of the other roses which were investigated. Negative relationships between the presence of starch and the organogenic capacity of callus have been recorded in other genera (George & Sherrington, 1984). Higher levels of respiration have been detected during the formation of shoots (Thorpe, 1978; Thorpe & Laishley, 1973) and it is possible that conversion of starch to sucrose provides energy for the organogenic response. Therefore, a physiological condition in which sugar is converted to starch may be incompatible with organogenesis. Furthermore, an increase in osmolarity, resulting from the conversion of starch to sucrose may lead to a greening of callus which initiates shoot formation (Thorpe & Meier, 1974a, b; De Fossard & Lee, 1974; Barg & Umiel, 1977). Adventitious shoots which formed from the roots of R. persica X xanthina were seen to arise in the region of the pericycle. Adventitious shoots of many species arise in his zone, perhaps as modified lateral roots (Peterson, 1975). Many roses characteristically exhibit the ability to produce root suckers in vivo, thus conditions leading to their formation in vitro merit further attention. Two species, R. laevigata and R. wichuraiana, did not produce adventitious shoots from either callus or roots, but did so from detached leaves. Induction of shoots from leaves has been exploited for the micropropagation of several species of ornamental plants (George & Sherrington, 1984). Techniques like the leaf-disc method used for induction of adventitious shoots in other genera of Rosaceae (James et al., 1984) might succeed in generating adventitious shoots in other roses.

Transplantation to soil is problematic in many roses, but particular difficulty was encountered with R. persica X xanthina. The use of Sorbarod plugs greatly facilitated transplantation of this hybrid. The particular merits of this system include the physical protection given by the plug to the rooted plantlet at the time of transplantation and, subsequently, the supply of water held within the cellulose matrix of the plugs. The plug is biodegradable and did not interfere with subsequent growth of the plants. The possibility that rooting, in vitro, is enhanced by improved aeration provided by the plug is presently under investigation.

Each of the methods used in the present study to induce adventitious shoots is of great interest to rose breeding as it allows the generation of somaclonal variants and the rapid segregation of chimeras resulting from treatment with colchicine or mutagens. The callus method has additional interest as a relevant approach in regenerating haploid roses from anther callus, such as that induced by Tabaeezadeh & Khosh-Khui (1981). It might also be used to regenerate somatic hybrids after protoplast fusion.


We thank J. Harkness for the gift of plant material and helpful discussions. D. Lloyd acknowledges the support of the Science and Engineering Research Council.