.

adc cDNA.

T. CAPELL, 0. LEPRI, P. HANG, L. BASSIE, N. NOGUEIRA, G. EL.-HUSSIENG, J, NEWMAN and P.

CHRLSTOU

Molecular Biotechnology Unit, John Innes Center, Colney Lane, Norwich NR7 ORT, UK

     Polyamines (PAs) comprise a class of low molecular weight compounds, which are implicated in a number of fundamental developmental and metabolic processes in plants. These include growth, development and stress responses. The simplest PA, putrescine (Put), is synthesized from the amino acids arginine (arg) or ornithine (orn) by the enzymes arginine decarboxylase (ADC) and ornithine decarboxylase (ODC), respectively. Additional steps convert Put into spermidine (Spd) and spermine (Spm). These steps are catalyzed by spermidine and spermine synthases, which add propylamino groups generated from Sadenosylmethionine (SAM) by SAM decarboxylase (SAMDC; Malmberg et a!. 1998). PAs also play a role in a wide range of biochemical processes in plants including DNA replication, transcription, protein synthesis, membrane stabilization and RNA and protein turnover (Evans and Malmberg 1989). These may have a direct bearing on the physiological processes that, in part, PAs are thought to mediate or regulate. Until recently, however, evidence for the direct role of PAs in these processes remained elusive. In part, this is a reflection of the large number of functions attributed to PAs and the difficulties involved in uncoupling them. Furthermore, most of the evidence for PA function has been correlative, in that various processes have been matched to differences in PA levels or distribution. Only in the last few years, with the use of transgenic plants and mutants in which specific enzymes in the PA pathway have been specifically over-expressed or inhibited, has it been possible to address the precise roles of PAs and to determine whether those roles are direct or indirect. The manipulation of PA levels in transgenic plants has been made possible through the cloning of genes involved in PA biosynthesis.

     We embarked on a systematic program aiming towards engineering the PA biosynthetic pathway in cereals using cloned genes. Our interests are two-fold: to unravel the role of PAs in morphogenesis and to create transgenic rice and other cereals with enhanced polyamine levels for nutritional improvement. We reported the recovery of fertile transgenic rice plants expressing the oat adc cDNA under the control of the CaMV 35S promoter (Capell et al. 1998) and demonstrated a correlation between PA accumulation and the ability of dedifferentiated tissue to undergo morphogenesis. In subsequent experiments, we analyzed transgenic rice callus lines and regenerating shoots expressing the oat adc cDNA under the control of the maize ubiquitin- 1 promoter. When we measured mRNA accumulation in callus lines containing pUbiADCs, the steady state RNA was detected in all transgenic lines and this was correlated to an increase in enzyme activity. In previous experiments we measured mRNA accumulation in transgenic rice callus containing the adc gene driven by the CaMV 35S promoter. Only a very small number of transgenic lines showed mRNA accumulation. The steady-state RNA in these lines (35S-adc) was only detectable in lines which were terminally dedifferentiated, i.e. not able to regenerate plants. Overexpression of the oat adc cDNA in transgenic rice driven by the strong constitutive maize ubiquitin- 1 promoter resulted in lines with altered levels of PAs. We have shown that levels of individual PA may vary significantly among the different lines and these depend on the developmental stage of the tissues. These results are shown in Figure 1. Dedifferentiated rice tissue undergoes morphogenesis when auxin is withdrawn from the culture medium. We measured PAs from callus lines that were induced to regenerate. Put and Spd levels were significantly higher than control values in all the lines and most of them were higher than values from samples growing in the fully dedifferentiated state (presence of auxin). Spm showed the reverse trend; values were significantly lower compared to controls and almost all lines had a reduction in endogenous levels compared to callus proliferating in the presence of auxin. This significant reduction in Spm in all lines compensates the increase of Put and Spd, resulting in only one third of the lines showing an increase in total PAs during the induction of morphogenesis. This may indicate that induction of morphogenesis requires an increase in ADC activity and accumulation of Put and Spd. Catabolic enzymes such as DAO and PAO are activated to keep spm at low levels possibly to compensate for the accumulation of Put and Spd in the pathway. After the onset of morphogenesis and the appearance of regenerating shoots on media without auxin, we observed a significant increase in Put in most shoots we analyzed. For Spd, 50% of the shoots showed a significant increase whereas the remaining half showed no variation compared to controls grown under the same conditions.. However, Spm levels showed a significant increase. Some of the lines had an increase in Spm levels (up to six-fold) compared to levels in the presence on auxin in fully dedifferentiated tissues. Our results suggest that Spm does not influence morphogenesis because all the lines were able to differentiate shoots. Total PA accumulation increased in all the lines only during shoot formation due to Put, Spd and Spm increases. During callus growth (dedifferentiated state) only one third of the lines showed variation (increase or decrease) in total PAs. Upon the onset of morphogenesis one third of the lines showed significant increases in total PAs compared to controls.