Results

4.1 Germplasm collection

Fifty six populations of A. paniculata were sampled (Fig. 4.1; Appendix I) from different geographical regions of tropical Asia. Out of 52 plant populations from India, 28 were from Kerala, 18 from Tamil Nadu, 2 each from Karnataka and Andhra Pradesh and 1 each from Assam and Maharashtra. There were 4 foreign populations, each from Indonesia, Malaysia, Sri Lanka and Thailand.

The two metapopulations collected from Western Ghats (Sirumalai, AP54) and Eastern Ghats (Nallamalai, AP56) consisted of groups of 9 and 2 local populations respectively. While each of the populations in the S (Sirumalai) group (S1 to S9) spatially separated by 1-5 km apart, the 2 populations of the N (Nallamalai) group (N1, N2) was from geographically isolated locations (allopatric) separated by at least 200 km. Each of the 11 local populations consisted of 2-8 subpopulations, sampled at 1 km apart.



4.2 Biology

4.2.1 Habitat

During the exploration surveys conducted at different parts of Kerala and Tamil Nadu, A. paniculata was found distributed in wild lands, as weed in cultivated lands and under cultivation in home gardens. It was quite common in fallow lands and also as undergrowth of scrubby or deciduous forests as well as in the midst of bushes and on the sides of the road/rail tracks. No particular association of this species with other plants was noticed. While the plants remained stunted or dried up during summer (March - September) in drier parts of Tamil Nadu, they showed prolific growth on the banks of the river Cauveri at Kollidam (Tamil Nadu) and in the wet regions of Kerala. The species was also collected from sea level in parts of Kerala and Tamil Nadu and up to 700 m msl at Alagar hills in Tamil Nadu. The wet and rich soil of Kerala, hot sandy beach at Kanyakumari and rocky scrubs of Alagar Hills and Jawadi Hills represented some of the extreme habitats from which the plants were collected.

4.2.2 Cytology

Microscopic examination of flower bud squashes in randomly selected plants from different phytogeographic regions revealed no change in ploidy level. Haploid chromosome number (n) of the species (based on three to five counts from each population) was uniformly 25 (Fig. 4.2).

4.2.3 Phenology of flowering and fruiting

Though flowering of A. paniculata is largely seasonal, initiation of flowering in the plant populations of southern India was noticed from December onwards. Plants growing in nature under favourable conditions produced flowers to varying degrees throughout the year. In drier parts of Tamil Nadu, flowering occurred mostly in February - April and fruiting and seed dispersal during March - June. Plants collected from Assam (AP29) showed early flowering (October - November) characteristics. The flowers were smaller, in spreading axillary and terminal racemes or panicles. The size of the flowers was invariant in comparison to the plant size. The flowers opened in the morning and continued till evening depending on the maturity of the flower bud. They usually dropped within 12 h. Percentage of fruit setting varied from 50-90, indicating that all the flowers did not form fruits. The capsules were linear-oblong, acute at both ends, about 1.8 cm long and 0.3–0.5 cm wide; seeds averaged from 8-10 per capsule and were sub-quadrate and yellowish brown.

4.2.4 Breeding system

Results of the manual selfing and bagging experiments indicated that
A. paniculata is self-compatible. Bagging experiment showed a mean fruit set value of 56.7% under conditions of limited chance for external stimuli
(e.g. wind) which could otherwise establish effective contact between stigma and anthers. Manual selfing, netting and open pollination resulted in about 72% fruit set. The bagging experiment was performed on the corridor of the lab to avoid falling of rainwater on the plant. Other experimental plants were reared in the field plot where there was enough natural wind to agitate the plants. The favourable effect of wind on pollination was further confirmed by indoor pollination experiments. When plants were kept in a room where the windows were closed and the fans switched off, there was only 10.8% fruit set. The results of pollination experiments are summarized in Table 4.1.



Table 4.1. Summary of the results of pollination experiments in A. paniculata


Thirty-two of the 34 emasculated flowers on crossing produced no seeds. But two of the flowers emasculated and pollinated by pollen from a different plant developed into a fruit each having six seeds in it. The emasculated and undusted flower buds did not get fruited as expected.

4.2.5 Seed dormancy and germination

Germination of freshly harvested seeds was negligible (<5%).>

In an experiment aimed at inducing germination of freshly harvested seeds, treatment with GA3 and Kinetin resulted in considerable increase in germination (20% and 24% respectively; Fig 4.4). Among the surfactants, 10% germination was recorded with Tween 80 treatment while CTAB or acid scarification process did not have any effect.



4.3 Analysis and estimation of intraspecific variations

4.3.1 Morphology

The polygraphic representation (Fig. 4.5a) of the quantitative morphological characters revealed considerable diversity (Fig. 4.5b). Phenotypic variations among the plants raised under identical conditions of growth were not uncommon (Fig. 4.6). They differed for plant height, branching, number of leaves and reproductive fitness. In general, population AP23 record maximum for plant height with a mean value of 81.36 cm and AP51, the lowest with a mean of 25.50 cm. Plants raised from seeds of all the populations except AP51 did not show branching up to 8-10 weeks after which they branched off with varying frequencies under uniform growing conditions. The population showing maximum number of branches was AP51 (mean 60.4) while AP53 showed the lowest (mean 12.0). Leaf number was significantly higher than others in AP51 with a mean value of 988.8 after 6 months of growth. Among the various characters indicating the reproductive fitness of the individual populations, AP18 and AP06 had maximum (mean=334.1) and minimum (mean=75.9) number of fruits respectively.


Fig. 4.3. Relationship between seed storage, seed moisture content and seed germination in A. paniculata




Fig. 4.4. The effect of various treatments on inducing germination in freshly collected seeds of A. paniculata


Regarding qualitative attributes, pigmentation in leaves and flowers was characteristic of certain populations. For example, AP34 showed pinkish green leaves towards their maturity (Fig 4.7). This became prominent when the plants were exposed directly to sunlight. Flowers were uniform in all populations except AP51 that showed relatively larger flowers with characteristic pigmentation and brownish spots on petals (Fig 4.8)

Fig. 4.5a. Polygraphic representation of selected quantitative morphological characters of A. paniculata



Fig. 4.5b. Polygraphs showing diversity in morphological characters among
15 populations of A. paniculata.

4.3.2 Andrographolide content

Concentration of andrographolide (Fig. 4.9) in the leaves varied from 0.73 (AP34) to 1.47% (AP36) with a mean value of 0.95 as evidenced from spectrophotometric analysis. The andrographolide yield was stable through three successive progenies (Fig. 4.10). The mean value of andrographolide estimated from each generation and the overall mean for the 3 generations are presented in Table 4.2. The compound extracted from the leaves and further analysed through HPLC in different populations confirmed results obtained with UV spectrophotometry. A comparative profile of the authentic sample and that extracted from AP36 is shown in Fig 4.11 and 4.12.



Table 4.2. The andrographolide yield in 15 populations of A. paniculata for the three successive generations.


Fig. 4.9. Chemical structure of andrographolide (C20H30O5; MW 350.44)



Fig. 4.10. The mean andrographolide content of three successive generations in different populations of A. paniculata. Seeds were collected from individual plants and were sown to raise the next germination. The error bars indicate average standard deviation (SD) values of 0.16.


4.3.3 Variation in pigment content

Since colour of the leaves showed variations in certain populations, quantitative analysis of chlorophylls, carotenoids and anthocyanins was done. Pigment content obtained with acetone extract of young leaves of 6 months old plants revealed considerable differences in the 15 populations (Table 4.3). Total chlorophyll content varied from 0.026 (AP24) to 0.058 (AP18) with a mean value of 0.039. Concentration of carotenoids was the lowest in AP24 (2.071) and highest in AP18 (5.393) with a mean value of 3.486. However, the values obtained for anthocyanin (A530) was highest the (1.961) for AP24 and lowest for (0.688) in AP06.



Table 4.3. Concentration of chlorophylls, carotenoids and anthocyanins obtained from different populations of A. paniculata. The value of each population represents average of 4-8 young plants. Acetone (80%) and methanol extracts of young leaves were used for the analysis. Chlorophylls and carotenoids are expressed in mg/ml and anthocyanins in absorbance (A530) values.

In order to study chance fluctuations in pigment composition at maturity, the leaves of the fruit bearing mature plants (8 month old) of selected populations were subjected to quantitative analysis (Table 4.4). The amounts of chlorophylls and carotenoids were decreased by 33.33% and 41.59% respectively while concentration of anthocyanins increased by 66.34% in selected populations of AP18, AP46, AP51 and AP52. The leaves of AP24 showed less concentration of anthocyanin at maturity compared to earlier stage and as against other populations which recorded increased concentrations.



Table 4.4. Concentration of chlorophylls, carotenoids and anthocyanins obtained from different populations of A. paniculata. The value of each population represents average of 4-8 mature plants. Chlorophylls and carotenoids are expressed in mg/ml and anthocyanins in absorbance (A530) values.

4.3.4 Chromosome number

All the 4 randomly chosen populations showed no variations in chromosome number. The number of chromosomes counted in anther squashes was n=25.


4.3.5 Isozyme variations

(i) ex situ populations

Eight enzyme systems (Fig. 4.13) provided 15 loci for the 15 populations of A. paniculata. Out of these, 1 locus was found only in one population (AP51) while 8 were invariably present in all the populations. Of the 8 loci, Per1, Sod2 and Sod3 were monomorphic to all the populations analysed. Unique alleles (specific to a population) were detected in the feral type AP51 at Est3 locus. Rare alleles (present at low frequency in more than one population) were present at the dimorphic Sod1 locus in two populations (AP36 and AP51). It was monomorphic in the remaining 13 populations. At the species level, all the 3 alleles (A, B, C) were not equally distributed (Table 4.5). There was maximum variation in Est1, Got1, Got2, Adh1 and Sod1, moderate variation in Sdh1, Mdh1 and Gdh2, and minimum variation in Est2, Est3, Gdh1, Per1, Per2, Sod2 and Sod3 loci.



Table 4.5 Overall allele frequencies at 15 loci in 8 different enzyme systems (EST, GOT, ADH, SDH, MDH, GDH, PER, SOD) in 15 populations of A. paniculata.

GENETIC VARIABILITY. The genetic variability measures of the 15 populations are presented in Table 4.6. The mean observed number of alleles per locus ranged from 1.20 to 1.57 (mean 1.41±0.11). The proportion of polymorphic loci (P) (0.99 criterion) varied from 0.13 to 0.53 (mean 0.36±0.12), the mean observed heterozygocity (Ho) from 0.200 to 0.571 (mean 0.406±0.11), the mean expected heterozygocity (He) from 0.114 to 0.327 (mean 0.233±0.07) and gene diversity (I) varied from 0.138 to 0.396 (mean 0.281±0.08).

GENETIC DIFFERENTIATION AND RELATIONSHIPS. The genetic identity (i) and distance (d) values (Nei, 1972) were calculated between 15 populations and presented in Table 4.7. The analysis revealed that genetic identity was highest (1.000) between AP06 vs. AP21, AP06 vs. AP22, AP21 vs. AP22 and AP24 vs. AP28. It was lowest (0.488) between AP51 vs. AP36.


Table 4.6 Descriptive statistics (over all loci) of 15 populations of A. paniculata. P, proportion of polymorphic loci (0.99 criterion); A, mean observed number of alleles per locus, I, Shannon's Information index (gene diversity), Ho, observed heterozygocity and He, expected heterozygocity.


The UPGMA dendrogram (Fig. 4.14) was a fair representation of the Nei’s genetic distances between the populations, which showed many clusters. AP51 obtained from a local nursery (Pazhakulam, Kerala) was found to be genetically more distant from all other populations and stood alone in the dendrogram. Remaining 14 populations clustered at various levels.


Fig. 4.14 Dendrogram based on UPGMA clustering of 15 populations of
A. paniculata using the genetic distances of Nei (1972) derived from isozyme data




Table 4.7 Genetic relationship among 15 populations of A. paniculata from different localities.

(ii) in situ populations

Out of six loci inferred from studying the 11 local populations of the two metapopulationss of A. paniculata from Sirumalai and Nallamalai region using EST, GOT and SOD enzyme systems (Fig. 4.15), three (Got1, Sod1 and Sod2) were monomorphic (Table 4.8). Of the remaining three loci, two (Est2 and Got2) showed less amount of variation while one (Est1) was highly polymorphic.

GENETIC VARIABILITY. While various genetic diversity estimates were more or less stable within each population (Table 4.9), they showed considerable variations between them. The mean polymorphism, P of each population varied from 0.00 to 0.50 (with an overall mean value of 0.270.09); the observed number of alleles, na from 1.00 - 1.50 (1.320.07); the Shannon's gene diversity, I from 0.00 - 0.35 (0.230.05); the Nei's heterozygocity, h varied from 0.00 - 0.25 (0.170.03). The study revealed higher level of variability in Sirumalai (P=0.37; A = 1.40; I = 0.28; h = 0.20) than in Nallamalai populations (P=0.16; A = 1.24; I = 0.17; h = 0.13). At the local population level, N1 and N2 from Nallamalai showed only 0.19 and 0.13 polymorphism respectively, while Sirumalai populations showed comparatively higher P value, which ranged from 0.25 to 0.50. All other estimates (A, I, or h) showed higher values in Nallamalai population comparing to that from the other region.



Table 4.8 Overall allele frequencies at 6 loci in the 11 local populations of the two metapopulations of A. paniculata from Sirumalai and Nallamalai region.


GENETIC DIFFERENTIATION AND RELATIONSHIPS. Nei's (1972) original measure of genetic identity (i) and distance (d) values were calculated and presented (Table 4.10). Higher genetic identity (>0.990) was observed between N1 vs N2 and S5 vs S6. The values ranged from 0.469 (S3 vs N1)


Table 4.9 Descriptive statistics (over all loci) based on isozyme analysis within and between 11 local populations of the two metapopulations of A. paniculata from Sirumalai and Nallamalai region. P, proportion of polymorphic loci (0.99 criterion); na, mean observed number of alleles per locus, I, Shannon's Information index (gene diversity) and h, Nei's expected heterozygocity.

Fig. 4.16 A dendrogram based on UPGMA clustering of 11 local populations of the two metapopulations of A. paniculata from Sirumalai and Nallamalai region using the genetic distances of Nei (1972) based on isozyme data



Table 4.10 Genetic relationship among 11 local populations of the two metapopulations of A. paniculata from Sirumalai and Nallamalai region.


to 0.996 (S5 vs S6). The dendrogram (Fig 4.16) constructed based on Nei's distance values clearly differentiated Sirumalai and Nallamalai populations so that they were distributed into 2 clusters of the dendrogram.

Results of F-statistics analyses of genetic differentiation in the two metapopulations are presented in Table 4.11. Regionally there is 100% deficit of heterozygotes and averaging 32% and 72% in the Nallamalai and Sirumalai metapopulations respectively. The degree of genetic differentiation (FST) averaged at the N region was 0.34 while at the S region it ranged from 0.00 to 0.36 with a mean value of 0.15. The theoretical gene flow (Nm) calculated from FST values at the population level ranged from 0 to 3.13. All the 11 populations studied recorded highest level of selfing theoretically possible at the sub population level. The selfing at the level of entire population varied greatly at Sirumalai, ranging from -1.00 to 0.29.



Table 4.11 F-statistics of local populations of the two metapopulations of
A. paniculata from Sirumalai and Nallamalai region.


4.3.6 RAPD assay

(i) ex situ populations

For RAPD assay, altogether 27 random primers were used. Of these, few failed to produce any amplification while in certain others amplified bands were less distinct and hence not considered. Of the 10 scorable primers (Table 4.12), number of bands produced per primer ranged from 9 (OPAW-11) to 1 (OPZ-09 & OPZ-15) and the products ranged in size from 310bp to 3500bp (Fig. 4.17).



Table 4.12 List of primers and their base sequence employed in PCR for detection of RAPD variation in 15 populations of A. paniculata

The proportion of polymorphic loci ranged from 0.00 to 1.00 with a mean of 0.81. The effective number of alleles (ne) ranged from 1.14 to 1.99 with a mean value of 1.62±0.22. Shannon’s information index (I) (Shannon and Weaver, 1949) showed a mean RAPD value of 0.55±0.12 for the 15 populations (Table 4.13). The Nei’s (1973) average heterozygocity (h) values showed considerable differences among populations (p=0.05) and the mean value was 0.37±0.10. Genetic identity measures were estimated (Table 4.14), the values of which ranging from 0.343 to 0.943. The mean value was found to be 0.659±0.12 with a mean standard error (SE) of 0.01.



Table 4.13 Summary of genetic variation statistics of 15 populations of A. paniculata for all loci obtained through RAPD assays. [ne = effective number of alleles, h = Nei's average heterozygocity, I = Shannon's Information index, P = proportion of polymorphic loci].

The UPGMA clustering of 15 populations of A. paniculata using RAPD data yielded 4 main groups: group-1, group-2, group-3 and group-4 (Fig 4.18). Group-1 consisted of mostly populations from Kerala: AP23, AP24, AP06 and AP46. The exception was AP28 collected from Tamil Nadu. The group-2 has 5 populations of which 2 (AP21 and AP34) are from Kerala, 2 (AP33 and AP36) from Tamil Nadu and 1 (AP18) from Maharashtra. The group-3 consisted of 3 populations: AP48, AP52 and AP53. The last group consisted of only 2 populations viz. AP22 and AP51.



Fig. 4.18 Dendrogram based on UPGMA clustering of 15 populations of
A. paniculata using the genetic distance of Nei (1972) derived from RAPD data.


Table 4.14 Genetic identity values from RAPD data obtained in 15 populations of A. paniculata


(ii) in situ populations

The 2 metapopulations (AP54 and AP56) of A. paniculata comprising of 11 local populations and 36 subpopulations were analysed using 35 random primers. Of these, banding pattern obtained with some primers was either not reproducible or less distinct and hence they were not considered for the genetic analysis. The 15 primers selected (Table 4.15) yielded 2 (OPB18) to 10 (OPX10) bands per primer with an average of 5.9 bands and the products ranged from 200 bb to 3800 bb (Fig. 4.19).



Table 4.15 List of primers and their base sequence employed in PCR for detection of RAPD variation within 2 metapopulations (AP54 and AP56) of A. paniculata


Table 4.16 Summary of genetic variation statistics for all loci from the 11 subpopulations of the 2 metapopulations of A. paniculata. [P = proportion of polymorphic loci, ne = effective number of alleles, h = Nei's average heterozygocity, I = Shannon's Information index].



GENETIC VARIABILITY AND DIFFERENTIATION. The effective number of alleles (ne) ranged from 1.25 to 1.73 (Table 4.16). Shannon’s information index (I) showed a mean value of 0.40 and 0.35 for Sirumalai and Nallamalai populations respectively. The average heterozygocity (h) ranged from 0.13 to 0.34 among the local populations. The proportion of polymorphic loci ranged from 0.40 to 0.73 with a mean of 0.600.05. Genetic identity measures were estimated (Table 4.17), the values of which ranged from 0.707 (S8 vs N1) to 0.974 (N1 vs N2). Higher identity (>0.970) was observed between N1 vs N2 and S1 and S4. The dendrogram based on genetic distance of Nei (1972) was able to distinguish Sirumalai and Nallamalai populations which appeared as two major clusters having 2-9 branches (Fig 4.20).



Fig. 4.20 Dendrogram based on UPGMA clustering of the 11 local populations of the 2 metapopulations (AP54 and AP56) of A. paniculata using the genetic distance of Nei (1972) derived from RAPD data.


Table 4.17 Genetic identity values from RAPD data obtained with 11 local populations of the 2 metapopulations of A. paniculata


4.4 Selection of desired mutants

AP36, the population which yielded highest concentration of andrographolide was subjected to chemical mutagenesis and possible selection of desirable variants. For the 16 different combinations of varying time (1 - 12 h) and concentrations of EMS (0.05 – 1.00%), the seed germination rates (Table 4.18) varied from 0 - 96%. Altogether 13 mutant lines (provisionally designated as CM) were obtained and the remaining 3 could not be considered since CM16 did not germinate at all while CM7 and CM12 though germinated, the seedlings did not survive. In all the 13 surviving lines, 8-10 plants were used for further characterization and yield trials.

The mean dry weight (g) of the mutants (whole plants) at the time of flowering varied from 0.30 to 0.35 with a mean value of 0.32±0.16. The content of andrographolide on dry weight basis recorded a maximum value of 1.53% with an overall mean of 1.26±0.16%. Highest concentration of andrographolide was recorded in CM1 and CM3. These values were 10.7% and 16.8% more than the values recorded for the untreated control respectively. This high concentration of the active principle found in CM1 and CM3 could not be correlated with their corresponding esterase profile since the esterase banding pattern of the control, CM1 and CM3 was identical. The values of dry matter production and andrographolide content recorded more or less the same through two generations (2nd & 3rd) raised by seed germination (Fig. 4.22).



Table 4.18 Mutagenesis of A. paniculata (AP36) seeds: details of duration of treatment vs. percentage of EMS (v/v), germination percentage of treated seeds and corresponding ids.

At the genetic level chemical (EMS) mutation did not affect the EST enzyme system (Fig. 4.23). Out of the 2 loci scored, Est2 was invariant in all the presumptive mutants while Est1 showed minor variation. The Est1 locus consisted of 1-3 alleles with an average of 1.77 alleles. CM3, CM5 and CM15 showed 2 alleles at this locus while CM13 showed 3 alleles. In all other mutant lines Est1 was monomorphic. Quantitative variation as evidenced from intensity of bands at Est1 locus in some mutants was also noticed.


Fig. 4.22 Mean biomass and andrographolide content of the presumptive chemical mutants of A. paniculata (AP36) from 2 successive generations (2nd & 3rd). The error bars indicate average standard error (SE) values.


Fig. 4.23 Esterase profile of 13 mutant lines of A. paniculata (AP36)