A Study on Comparative Fertility Restoration in A2 and A4 Cytoplasms and Its Implication in Breeding Hybrid Pigeonpea [Cajanus cajan (L.) Millspaugh]

Exploitation of hybrid vigour has been visualized as the most efficient option for increasing productivity in pigeonpea [Cajanus cajan (L.) Millspaugh]. Cytoplasms from various wild relatives of pigeonpea have been transferred to develop CMS lines in the background of cultivated pigeonpea. However, A2 (Cajanus scarabaeoides) and A4 (Cajanus cajanifolius) cytoplasms have been utilized most frequently. In order to study fertility restoration efficiency in F1 hybrids having either A2 or A4 cytoplasms, an experiment was conducted at the Indian Institute of Pulses Research (IIPR), Kanpur during 2008-2012. Four CMS lines namely Hy4A, H28A (each with A2 cytoplasm), ICP 2039A and ICP 2043A (both with A4 cytoplasm) were crossed with ten genotypes/restorers of long duration pigeonpea for two years. The F1 hybrids so-obtained were assessed in the succeeding years for pollen fertility and pod setting. All the pollinators except IPA 203 restored fertility in F1 hybrids derived from ICP 2039A and ICP 2043A (both having A4 cytoplasm). However, none of the restorers were effective in restoring fertility in hybrids derived from Hy4A and H28A (each with A2 cytoplasm). This could be ascribed to undesirable linkage drag still present in these two CMS lines having A2 cytoplasm. The F2 progenies derived from 4 hybrids (ICP 2039A × NA-1, ICP 2039A × Bahar, ICP 2043A × NA-1 and ICP 2043A × Bahar) segregated approximately into 3 fertile: 1 sterile plants. However, 2 F2 progenies having Pusa 9 as the restorer revealed approximately 15 fertile:1 sterile ratio. Thus monogenic and digenic duplicate gene action with complete dominance for fertility restoration was observed in F1 hybrids derived from CMS lines having A4 cytoplasm. F3 progenies from individual F2 plants of these crosses also confirmed the same pattern of fertility restoration. This study indicated that CMS lines based on A4 cytoplasm would be more desirable as these might have more number of restorers compared to those having A2 cytoplasm. Present address: ICAR Research Complex for Eastern Region, Research Centre for Makhana, Darbhanga, India. A. K. Choudhary, I. P. Singh


Introduction
Heterosis breeding was resorted to improve productivity of pigeonpea which has been static for the last three decades the world over [1].Pigeonpea fulfils several pre-requisites including higher outcrossing percentage for exploitation of hybrid vigour.Several cytoplasmic nuclear male sterility (CMS) systems are available in pigeonpea.However, CMS lines derived from Cajanus scarabaeoides [2] (A 2 cytoplasm) and C. cajanifolius [3] (A 4 cytoplasm) have been widely utilized to develop commercial hybrids.Despite release of a few hybrids from both sources, hybrids did not gain ground on farmers' fields due to several factors including partial fertility restoration and high genotype-environment interaction [4].
According to De [5], C. cajanifolius resembles cultivated types in most morphological traits.The CMS lines containing A 4 cytoplasm have been reported to be highly stable across environments and years without showing any morphological deformity [6].However, a comparative picture of hybrids derived from both A 2 and A 4 cytoplasm is scanty and also not well-documented.The present study reports a comparative assessment of fertility restoration in hybrids containing individually either A 2 or A 4 cytoplasm and their significance in breeding hybrid pigeonpea.In order to take advantage of this CMS hybrid technology, it is essential to breed high-yielding hybrids based on diverse genetic backgrounds.To achieve this, breeding of promising hybrid parents and knowledge of the inheritance of fertility restoration are also essential [7].Therefore, in addition to the F 1 , F 2 and F 3 generations were also generated from A 4 CMS lines to determine the nature of gene action in the F 1 generation, the segregation pattern in F 2 generation and its confirmation through F 3 generation in pigeonpea.
During the first year of experiment (2008), the two A 4 CMS lines (ICP 2039A and ICP 2043A) were crossed with all the 10 restorers.F 1 seeds were harvested separately, and grown during the next cropping season.All the F 1 plants were put under nylon net to observe pod setting.Pollen fertility reaction was also assessed with 2% acetocarmine.The same set of crosses was made again to observe the stability of fertility restoration in the ensuing season.Besides, the other two A 2 CMS lines (H28A and Hy4A) were also crossed with the same set of 10 testers to observe differences (if any) for fertility restoration.During the cropping season 2010, all the 20 F 1 's and 20 F 2 's derived from CMS lines ICP 2039A and ICP 2043A (A 4 cytoplasm) were grown in addition to 20 F 1 's descended from the two CMS lines containing A 2 cytoplasm.Pollen fertility was again assayed by the same procedures.Data were also recorded for segregation pattern on fertility restoration in F 2 generation for six crosses (ICP 2039A × NA-1, ICP 2039A × Bahar, ICP 2039A × Pusa 9, ICP 2043A × NA-1, ICP 2043A × Bahar and ICP 2043A × Pusa 9).All the F 2 plants were bagged under nylon net to observe pod setting.20 crosses involving the two CMS lines (H28A and Hy4A) with the same set of 10 testers were repeated again to observe stability of performance in the next generation.F 3 seeds from randomly chosen 10 F 2 plants from all the six crosses were grown during the year 2011-12.Data were recorded for the number of fertile and sterile plants in selected F 3 families.In addition to this, the 20 F 1 's of the previous season were also grown to observe the breeding behaviour.The same procedure was followed to observe fertility reaction in F 1 hybrids derived from CMS lines having A 2 cytoplasm.For determining pollen fertility in each generation, five fully developed floral buds were taken randomly from each plant and the anthers were squashed in 2% aceto-carmine solution.The pollen fertility of each plant was studied under light microscope.The densely stained pollen grains were considered as fertile, while the empty or partially stained pollen grains were assessed as sterile.The chi-square test was applied for the goodness of fit to different expected ratios in F 2 and F 3 generations.The entire experimentation was performed during 2008-12 at IIPR, Kanpur.

Results
Pollens of all F 1 hybrids (having A 4 cytoplasm) except "ICP 2039A × IPA 203 and ICP 2043A × IPA 203 were observed densely stained with 2% acetocarmine, and hence showed fertile pollen reaction during the year 2009.All such pollen fertile F 1 hybrids were observed to have normal pod setting under nylon net (Table 2(a)).The same fertility reaction was noticed in the next year also (2010), confirming the results of previous year.This indicated that all pollinators (except IPA 203) efficiently restored fertility in F 1 hybrids, and thus these crosses could be assessed for yield and other attributes.When hybrids containing A 2 cytoplasm were analysed for fertility reaction, pollens did not take stain at all.None of the hybrids set pods under nylon net (Table 2(b)), revealing that none of the 10 pollinators was able to restore fertility in any one of F 1 hybrids during the year 2010.The same set of F 1 hybrids having A 2 cytoplasm was assessed further for fertility reaction during 2011.It again showed the same results, showing consistency of performance for fertility restoration.Thus it was obvious that a total of 9 pollinators (out of 10) were able to restore fertility in F 1 hybrids having A 4 cytoplasm; however, none

Discussion
It is known that CMS system is a maternally inherited trait governed by specific (mitochondrial) genes which do not affect otherwise other properties of the plant [9].The fertility restorer (Rf or Fr) genes in the nucleus suppress the expression of male-sterile phenotype, leading to commercial exploitation of the CMS system for the production of hybrid seeds.Commercially exploitable CMS system has not been found in cultivated pigeonpea.Therefore, various wild relatives have been utilized to develop CMS system.The CMS system containing A 2 cytoplasm appears to reduce reproductive fitness of plants due to presence of several undesirable wild genes from C. scarabaeoides.This has been empirically observed in GTH-1, the first CMS based hybrid in pigeonpea.On the other hand, C. cajanifolius, which is the immediate progenitor of pigeonpea, resembles cultivated types in most morphological and agronomic traits [5].The male-sterile lines derived from A 4 cytoplasm are the best as they do not show morphological deformity and other fitness-reducing traits across environments and years [10].All these accounted for discrepancies in fertility restoration in F 1 hybrids containing A 2 and A 4 cytoplasm.
The knowledge of inheritance pattern of fertility restoration is indispensable for the transfer of restorer genes from one genotype to another.In the present study, it was observed that two restorers "NA-1" and "Bahar" showed monogenic inheritance (3:1) when crossed each with ICP 2039A and ICP 2043A, while "Pusa 9" revealed digenic inheritance of fertility restoration with both the CMS lines.The similar pattern of fertility restoration has also been reported in three diverse early maturing lines of pigeonpea [11].Variable restoration patterns among a common set of restorer lines (male parents) within a single cytoplasmic source of pigeonpea has been reported by Nadarajan et al. [12].In another study, it has been observed that the fertility restoration in A 4 CMS lines of pigeonpea may be controlled by either one or two fertility-restoring genes [13].In the present study, one fertility restorer line "Pusa 9" produced different results compared to "NA-1" and "Bahar" when crossed with the same set of A 4 CMS lines.The variable expression of fertility restoration can be attributed to different genetic backgrounds of the F 1 plants, arising from male parents of different genetic constitution.Alternatively, differences observed in segregation patterns also could be due to the presence of some modifier genes that influence the process of penetrance and expressivity of the fertility-restoring genes [14].On the contrary, the same pollinator (restorer) may also produce variable results if crossed with different A 4 CMS lines [7].

Conclusions
Pigeonpea is an important source of dietary protein especially for vegetarians of India as well as East Africa.Despite its global importance, the increase in its productivity has not been significant as it still possesses several wild traits including its perennial nature.Hybrid technology has been envisaged as one of the technological interventions to realise quantum jump in its productivity.For successful exploitation of hybrid vigour, CMS lines from various wild relatives have been developed.However, CMS lines containing A 2 (C.scarabaeoides) and A 4 (C.cajanifolius) cytoplasms have been widely used to develop high-yielding stable hybrids of pigeonpea.Although some CMS based hybrids have been made available for cultivation, these are yet to find commercial worth at farmers' fields.In this paper, we have examined a relative worth of A 2 and A 4 CMS lines for producing specific cross combinations and genetics of fertility restoration in A 4 CMS lines.The results indicated that A 4 CMS lines could provide larger number of cross combinations that could be assessed across years and locations as more number of pollinators could restore fertility in F 1 hybrids.As such, A 4 cytoplasm (derived from immediate progenitor of pigeonpea, C. cajanifolius) had displayed unconditional advantages over A 2 cytoplasm.In the present study, fertility restoration in A 4 CMS lines of pigeonpea was found to be cross-specific and influenced by the nuclear background of fertility-restoring lines.In 4 crosses (ICP 2039A × NA-1, ICP 2039A × Bahar, ICP 2043A × NA-1 and ICP 2043A × Bahar), fertility restoration was governed by a single dominant gene; while in 2 crosses (ICP 2039 × Pusa 9 and ICP 2043 × Pusa 9), it was controlled by two duplicate dominant genes.The differential behaviour of the two A 4 CMS lines (ICP 2039A and ICP 2043A) in crosses with "NA-1" and/or "Bahar" and "Pusa 9" could be ascribed to the interactions of different nuclear genes of the restorer male parents.

Table 2 .
(a) Fertility restoration in A 4 cytoplasm based pigeonpea hybrids; (b) Fertility restoration in A 2 cytoplasm based pigeonpea hybrids.

Table 3 .
Segregation pattern for male-sterile and male-fertile plants in F 1 , F 2 & F 3 generation of crosses involving A 4 cytoplasm.