Update on the genetics of flowering

Pisum Genetics

2007—Volume 39

Review

Update on the genetics of flowering

Weller, J.L. School of Plant Sci., Univ. of Tasmania, .Hobart, Tasmania, Australia

The biological mechanisms controlling flowering and photoperiod responsiveness have been of interest
to plant biologists for nearly a century, and the genetic control of flowering in pea has been under
investigation for a similar length of time. For a period in the 1970s, peas held a prominent place as a model
species for physiological genetics of flowering, due largely to the efforts of Ian Murfet. In a career spanning
more than forty years Ian, together with his colleagues and students, identified more than a dozen major
flowering loci through analysis both of natural variants and induced mutants. He also mapped many of
these loci, and used various physiological and genetic approaches to define their functions and
interactions.

Most of this early work preceded the molecular era, and until recently the molecular nature of the pea
flowering loci has remained largely unexplored. However, over the last decade work in arabidopsis has
given major insights into the genes and genetic mechanisms controlling plant responses to photoperiod
and temperature, flower development, light perception and endogenous rhythms (3, 18, 19, 24, 41). Partly
as a result, there have also been significant advances in the molecular biology of flowering in several other
model species, including rice, barley, wheat and tomato (9, 14, 23). This information, together with the
availability of extensive sequence databases in a number of model legumes (38) and the well-documented
synteny between pea and medicago (2, 20) has opened up a number of avenues for molecular analysis of
flowering in pea. Over the last few years work on flowering in pea has resumed in Hobart, with the
isolation of numerous flowering gene homologues and new flowering mutants (15, 16).

In this update I will summarize some of these more recent developments after first providing some
background information. Numerous reviews of earlier work are available (30, 31, 37, 51) and can be
consulted for further information.

1. Flowering loci

(a) LATE FLOWERING (Lf): A major flowering gene linked to A (now linkage group II) had been

observed by many workers before the Lf locus was definitively described as one of four major loci
contributing to the genetic variation for flowering time among existing pea cultivars (27, 29). Numerous
induced lf mutants are now also known, and all flower earlier than their respective progenitor lines (42).
The most severe mutants (Murfet's lf-a class) flower as early as node 6, but most cultivars appear to carry
an intermediate allele in the lf or Lf class and it seems likely that the ancestral form is represented by the
Lf-d class (29, 42). Lf is considered to govern the plant's "inherent lateness", because allelic variation at Lf
does not appear to interfere with the plant's ability to respond to photoperiod.

The Lf locus is notable as the first of the classical pea flowering loci to be identified at the molecular
level. One of three pea homologs of arabidopsis TFL1 (TFL1c) was identified as a candidate gene for Lf
based on its map position, and several lf-a class mutants were shown to have large deletions or amino acid
substitutions in TFL1c consistent with a complete loss of function (11). The isolation of an additional
EMS mutant (lf-22) carrying a nonsense mutation has provided further support for this conclusion (V.
Hecht, J. Weller unpubl.). In arabidopsis, mutations in TFL1 confer both early flowering and a
conversion of the indeterminate primary inflorescence to a flower (8). Although the primary inflorescence
of severe lf mutants remains indeterminate, a small function of Lf in determinacy is apparent in early
secondary inflorescences, which in lf mutants tend to terminate in an abnormal flower, instead of the
normal indeterminate stub.

Although the deletion and nonsense mutants clearly demonstrate that Lf is TFL1c, variation in
flowering time attributed to allelic variation at Lf is not always associated with mutation in the Lf coding
sequence. For example, the isolines WL1771/1770/1769 (Lf-d, Lf and lf, respectively) have no

Pisum Genetics

2007—Volume 39

Review

polymorphism within the coding region or introns of Lf but Lf expression does correlate with flowering
across this series (11). However, the possibility that these lines may carry mutations in the Lf promoter
region has not yet been excluded.

Despite the importance of Lf for flowering time, it is not known how Lf participates in mechanisms
controlling flower transition. Grafting experiments suggest Lf acts in the shoot apex, as its effects are not
graft-transmissible (26). Preliminary results suggest that expression of Lf occurs throughout the plant and
does not show any marked developmental or environmental regulation (11; B. Wenden, V. Hecht, C.
Knowles, unpublished), and it will be interesting to see if this is supported by more detailed studies.

(b) Photoperiod-specific repressors of flowering: As in the case of Lf, the effects of allelic variation at
the STERILE NODES (Sn) locus have probably been under study for more than a century, but it was
only with the use of controlled-photoperiod conditions that the existence of this locus could be clearly
demonstrated (4, 27). Mutant sn plants flower early under both LD and SD but, in contrast to lf mutants,
are unable to respond to photoperiod and when grown in SD display the short reproductive phase and
rapid reproductive development typical of WT plants in LD. Mutants at two other loci, DIE
NEUTRALIS (Dne) and PHOTOPERIOD (Ppd) have early-flowering, day-neutral phenotypes similar to
sn (1, 21, 44). The origin of the original Sn/sn allelic difference is obscure, but three additional induced sn
alleles have now been identified(1; S.E. Jones, J. Vander Schoor, J. Weller unpublished).

Comparisons with the arabidopsis system suggest that the majority of early-flowering photoperiod-
insensitive mutants have primary defects in maintenance of the circadian clock. We are currently
examining the expression of circadian clock gene homologues in sn, dne and ppd mutants and find that all
three show defects in rhythmic gene expression under light/dark cycles and constant conditions (V. Hecht,
L.C. Liew, unpublished). In parallel with these physiological studies, we are refining map positions for all
three loci (1, 21, 48) and examining relationships with candidate circadian clock genes in corresponding
regions.

Throughout the 1960s and 1970s, a variety of different grafting experiments was used to explore how
the Sn locus might influence the transmission of graft-mobile flowering signals. The majority of grafts
were performed epicotyl to epicotyl, effectively examining transmission from roots/cotyledons to shoot
apex. A small delay in flowering of sn scions induced by WT stocks was interpreted to suggest that sn
impaired production of a mobile inhibitor (4, 27). Similar conclusions were later reached for dne and ppd
mutants (21, 44). However, in all experiments of this type a strong promotion of WT scions by early
mutant stocks was also observed making it equally plausible that the mutants possess elevated levels of a
mobile floral stimulus.

More recently, we have been re-examining this question in grafts with leafy stocks possessing 4-5 true
foliage leaves. In this system the influence of the cotyledons has declined and flowering of the scion is
primarily determined by the influence of the stock leaves. We observe substantial promotion of flowering
in WT scions by early mutant stocks, but no significant inhibitory effect of WT stocks on early mutant
scions (L.C. Liew and J. Weller, unpublished), consistent with the view that genes of this nature
predominantly act through regulation of a mobile flowering stimulus.

(c) HIGH RESPONSE (Hr): Hr was another of the four major loci initially characterized by Ian Murfet
(28). Like Sn, Dne and Ppd, the dominant Hr allele inhibits flowering mainly under SD. In an otherwise
WT background, this inhibition may be so strong as to confer a near-obligate requirement for long days.
This suggests that Hr can be viewed as a photoperiod response gene, and evidence from grafting
experiments suggest that leafy hr stocks can strongly promote flowering in Hr scions and that Hr may act
through the same mobile signal as Sn (34, 35). One possibility is therefore that Hr, like Sn, Dne and Ppd,
may have defects in rhythmic gene expression and a primary role in the photoperiod response pathway.

However, another possibility is that Hr may be analogous to FRIGIDA (FRI) and FLOWERING
LOCUS C (FLC) in arabidopsis. These loci are typically discussed as mediators of the vernalization

Pisum Genetics

2007—Volume 39

Review

response (41) and are not generally considered as part of the arabidopsis photoperiod pathway, although
they do influence the photoperiod response through a dramatic delay in flowering under SD. Genes in the
FLC clade have not yet been conclusively identified in pea or in any model legume despite the existence of
extensive EST and genomic databases. On the other hand, FRI homologues are known from a range of
species, and orthologues of both FRI and FRIGIDA-LIKE1 (FRL1) are present in medicago (15).
Updated medicago mapping data suggests that MtFRI and MtFRLa both map on chromosome 3, and
that position of MtFRI corresponds to the approximate location of Hr in pea LGIII (J. Weller,
unpublished data).

(d) Photoperiod-specific promoters of flowering: In arabidopsis, several mutants with a LD-specific
late-flowering phenotype were among the first flowering mutants isolated (22) and the corresponding
genes have all turned out to be important components of the photoperiod response mechanism (14, 18, 33).
We have therefore been particularly interested to find mutants of this type in pea. The phytochrome A
(phyA) mutants were first identified by their defective de-etiolation responses to far-red light, and
subsequently shown to flower late in LD with additional phenotypes that are essentially a phenocopy of
WT plants grown in SD (increased basal branching, delayed senescence) (50). Mutations in the PHYA
gene are necessary for the promotion of flowering in response to photoperiod extensions rich in red light
but have little effect on the response to blue light (32, 52). A dominant, hypermorphic phyA mutant,
phyA-3D, was also identified in seedling screens, with an early flowering phenotype in SD similar to the sn,
dne and ppd mutants (53).

More recently, we have isolated a number of other mutants with phenotypes similar to phyA (16). Like
phyA mutants, late bloomer 1 (late1) mutants flower late in LD and have the general appearance of SD-
grown WT plants. Mutant late1 plants also have defects in rhythmic expression of circadian clock genes,
suggesting that Late1 may have a primary role in clock function (16). Consistent with these roles, Late1 is
the pea ortholog of arabidopsis GIGANTEA (16), which has a central role in circadian clock function and
additional, independent effects on photoperiodic flowering (25). The late flowering phenotype of late1
mutants is rescued by grafting to leafy WT stocks in LD, indicating that LATE1, like Sn, Dne and Ppd,
acts through regulation of a mobile flowering stimulus (16).The LATE BLOOMER 2 locus has a mutant
phenotype similar to phyA and late1, but has yet to be further characterized.

(e) Gigas: The Gigas locus is currently defined by two recessive mutant alleles. In SD both gigas
mutants flower later (10 to 20 nodes) than their respective WT, but otherwise show little phenotypic
difference. In LD, gigas mutants also show delayed flowering, but have a striking phenotype distinct from
photoperiod mutants phyA and late1. Mutant gigas plants develop normally until around the time WT
plants flower, and then undergo a striking "vegetative shutdown" in which internodes become shorter and
thinner, and the axillary buds buds at these nodes are released. The main shoot may eventually produce
one or two flowering nodes, but in other cases flowers may only be formed on lateral branches, and in the
strongest expression of the phenotype, the plants may never flower (7, 43). Expression of the gigas
phenotype is also influenced by light quality and temperature, and mutants are more likely to flower in
response to supplementation with light of low R:FR ratio (1; J. Weller, unpublished), under higher
irradiances (43), or at lower ambient temperatures (J. Weller and J. Vander Schoor, unpublished).

Grafting of gigas mutant scions to WT stocks can result in a significant promotion of flowering (7, 43),
leading to the suggestion that Gigas is involved in production of a mobile floral stimulus. However, as the
LD phenotype of gigas is distinct from phyA and late1, it seems likely the Gigas-dependent mobile signal
does not mediate all aspects of the photoperiod response but is limited to the initiation of flowering. One
possibility is that the Sn, Dne, Ppd and Late1 genes all act through Gigas to regulate the same mobile
flowering stimulus, but it is possible that they may also affect other, Gigas-independent, mobile signals.
New grafting experiments are underway to examine this question.

Pisum Genetics

2007—Volume 39

Review

Comparative mapping in pea and medicago locates Gigas near a cluster of medicago genes similar to
the arabidopsis FT gene (16). The arabidopsis FT gene has an important role in integration of flowering
signals and mobile signalling from leaf to apex, and in light of grafting results the possibility that Gigas
corresponds to an FT-like (FTL) gene in pea remains attractive. Isolation of pea FTL genes is in progress.

(f) Other flowering loci: Two other genes with a primary role in light responses also have effects on

flowering. Mutations in the PhyB gene confer early flowering phenotype that is primarily apparent in SD.
However, a null phyB mutation is epistatic to both phyA and late1 mutants in LD, showing that phyB can
act to delay flowering in both LD and SD (16, 52). This also suggests that PhyA and Late1 genes promote
flowering in LD by opposing a PhyB-dependent inhibition. Unlike sn, dne and ppd mutants, phyB
mutations only affect the node of flower initiation and do not markedly alter other pleiotropic aspects of
the photoperiod response.

The light-independent photomorphogenesis 1 (lip1) mutant was isolated as a spontaneous mutant
showing a constitutively de-etiolated appearance even when grown in complete darkness (12). In this
respect lip1 is similar to the COP/DET/FUS mutants of arabidopsis, and has been shown to carry a
complex duplication/ rearrangement in the pea COP1 ortholog (40). The original lip1 mutant arose
spontaneously in a genetic background (nominally cv. Alaska) carrying an sn mutation, which masked
any effects of lip1 on flowering. However, after selection away from sn and introgression into the cv.
Torsdag background it has become evident that lip1 mutants are somewhat early flowering in SD and
show a reduced photoperiod response similar to sn, dne, ppd and the phyA-3D mutant (J. Weller
unpublished).

Among a wide range of flowering mutants obtained from recent screens, we have identified two other
new LATE BLOOMER loci, Late3 and Late4. The late3 and late4 mutants have a novel flowering
phenotype characterised by extremely late flowering and a delay in the compound leaf transition under
both SD and LD (J. Weller and J. Vander Schoor, unpublished). Mutants do not commence flowering
until after node 35 and thereafter abort flower initials, fail to set pods, and occasionally show vegetative
reversion. Some pods do eventually form at later reproductive nodes, but show very weak growth and
yield few seeds. The late3 and late4 mutants continue to grow almost indefinitely in a cool environment if
free from disease, and exhibit a massively extended reproductive phase. Although nearly sterile, they do
not display the vegetative shutdown seen in gigas mutants, flower-sterile mutants or WT plants from
which flowers have been removed. Nor do they exhibit basal branching or other SD characters in LD like
phyA, late1 and late2 mutants. Instead, late3 and late4 produce strong aerial lateral branches later in
development. Preliminary evidence indicates that the late3 and late4 phenotypes are neither rescuable nor
transmissible through grafting.

Several other flowering loci, including E, Lw and Dm, have been described in various earlier reports
and reviews (31, 37, 49), but no new information about these loci has become available since the last
review (37). A role for the Aero locus in flowering has also recently been reported (47). However, with the
exception of early work on E (27, 31), the relationship of these loci with other flowering genes has not been
explored.

2. Inflorescence identity loci

The pea inflorescence is a compound raceme, and its development has been discussed in several reviews
(5, 39). A number of mutants affecting inflorescence and floral development have now been characterized
at the molecular level. The unifoliata (uni), proliferating inflorescence meristem (pim) and stamina
pistilloida (stp) mutants predominantly affect the floral meristem, and the Uni, Pim and Stp genes
correspond to the arabidopsis LFY, AP1 and UFO genes, respectively (6, 17, 45, 46). Although all three
mutants also have additional defects in development of the secondary inflorescence, they undergo a clear
transition to flowering at a similar node to WT and produce peduncles clearly distinct from vegetative
shoots. They therefore seem able to correctly specify both primary and early secondary inflorescence

Pisum Genetics

2007—Volume 39

Review

development. This implies the existence of additional, earlier-acting genes that also participate in
secondary inflorescence development, and several such loci are known.

The Det locus has a negative role in secondary inflorescence development, acting to prevent expression
of the secondary inflorescence program in the primary inflorescence meristem. This role is analogous to
that of arabidopsis TFL1, and Det is now known to encode another of the three TFL1 homologs (TFL1a)
in pea (11). The genetic interactions and molecular consequences of det mutations have yet to be explored.

Three other loci have a positive role in secondary inflorescence development. The VEGETATIVE1
(Veg1 ) locus (formerly VEGETATIVE; Veg) is represented by a single mutant allele. Homozygous
mutant plants never produce flowers, and must be maintained through the heterozygote (13). Despite
their failure to flower, veg1 mutant plants grown in LD clearly undergo a vegetative shutdown similar to
gigas (7, 36), suggesting that the photoperiod response mechanism is intact but the conversion of
vegetative to primary inflorescence meristem is blocked. Comparative mapping in pea and medicago has
located Veg1 near two MADS box genes that are homologues of arabidopsis FRUITFULL and

SEPALLATA1 (15).

A second locus VEGETATIVE2 (Veg2) has yet to be described in a primary research paper, but
descriptions of two mutant alleles are available (30, 31). The stronger of the two alleles confers a non-
flowering phenotype similar to veg1. However, a weaker allele, veg2-2, displays an unique phenotype that
reveals the role of this gene in secondary inflorescence development. Commencing at the node of flower
initiation in WT, axillary branches of veg2-2 plants are released, and produce a series of axillary structures
varying more-or-less continuously from normal lateral branches at lower nodes to normal secondary
inflorescences and flowers at higher nodes. In intermediate lateral structures, flowers may be produced
directly from nodes as in a normal secondary inflorescence, but there is a failure to suppress leaf formation
and to terminate apical growth. Recent data confirm that Veg2 is located on the bottom half of linkage
group I (J. Weller and I. Murfet, unpublished).

We recently identified a third locus in this group, LATE BLOOMER 5 (Late5). The single known late5
mutant allele shows similarities to the weak veg2-2 allele, resulting in late flowering, partial loss of
secondary inflorescence identity, and floral abnormalities. Although not allelic with Veg2, preliminary
results also locate Late5 to the bottom of group I (J. Weller and S. Davidson, unpublished). Interestingly,
the corresponding region in medicago includes homologues of the arabidopsis genes FD and SVP (J.
Weller and V. Hecht, unpublished), and we are currently examining the relative map positions and
relationships of these genes.

3. Isolation, mapping and expression analysis of flowering genes

We previously reported the isolation and mapping of many different pea homologues of arabidopsis
flowering-related genes (15). This work is continuing and additional flowering related gene homologs
identified, isolated and mapped in pea and/or medicago include PHYE, FRI, SVPb, PRR3/7, PRR5/9,
TIC, FHY3, SHP, STK, SPA1/2 and CDF1/2, LUX, FTLd/e and FD (V. Hecht, L.C. Liew, C. Knowles
and J. Weller, unpublished data). Where relevant we are now examining the transcriptional regulation of
many of these genes in studies of circadian rhythms, light and temperature responses, mobile signalling
and inflorescence development.

Of particular relevance to photoperiodic flowering are the CONSTANS (CO) and FLOWERING
LOCUS T (FT) gene families, both of which appear to have undergone differential expansion compared to
their arabidopsis counterparts (15, 16). New comparative mapping data suggest locations for the four pea
Group I CO-like (COL) genes (COLa-COLd) in LGV, LGII, LGIII and LGIV, respectively. Interestingly
COLa, the most similar pea gene to AtCO, shows a different diurnal expression pattern than AtCO, and is
not regulated by Late1, suggesting that it may have a different role than AtCO (16).

Similarly, the FT family in arabidopsis contains two genes (FT and TSF) but there appear to be at
least five in pea and medicago (V Hecht, J Weller unpublished). By inference from medicago the pea genes
are expected to be located in two clusters, in the middle and bottom of LGV. At least one of these genes is

Pisum Genetics

2007—Volume 39

Review

specifically expressed in expanded leaves under LD (C Knowles, V Hecht unpublished) and this expression
is greatly reduced in the late1 mutant (16). This suggests that pea FTL genes may have broadly conserved
roles, and we are now carrying out detailed expression studies of FTL genes in a wide range of different
conditions and mutant backgrounds.

4. Conclusions

Our results so far have already indicated that a number of changes are necessary to previous working
models for flowering in pea. Recent comparative studies in a range of species suggest a broad conservation
of flowering mechanisms (9, 14, 18, 41), and we have found it useful to move to a comparative model based
generally on arabidopsis. In arabidopsis, the FT protein acts as a mobile flower-promoting signal that
integrates light, daylength, circadian clock, ambient temperature and vernalization inputs (3, 10, 33), and
the molecular phenotypes of most pea mutants seem to fit at least generally with such a model. However,
that is not to say that are no significant differences between the pea and arabidopsis flowering systems. In
fact, it is already clear that there are several points of difference—concerning, for example, the roles of CO-
and FT-like genes, the pleiotropic nature of the photoperiod response, the specification of the secondary
inflorescence, and the nature of the vernalization mechanism.

It seems likely that mapping, expression studies and physiological analyses will soon help to identify
the molecular basis for many of the mutants collected in Hobart over the past forty years. We hope that
this will help us to understand the mechanisms regulating flowering in pea, and in particular, to give us
insight into those mechanisms that are divergent or perhaps even unique in pea. This should in turn yield
valuable information for the genetic analysis of flowering in related legumes such as lentil, medicago,
clover and chickpea.

Acknowledgments: I thank my colleagues Valerie Hecht, Jackie Vander Schoor, Claire Knowles, Lim Chee Liew and
Sandra Davidson for permission to refer to their unpublished results, and John Ross and Valerie Hecht for their
comments on the manuscript. Work on flowering and pea genetics in Hobart has been funded by the Australian
Research Council for many years.

1. Arumingtyas, E.L. and Murfet, I.C. 1994. J. Hered. 85: 12-17.

2. Aubert, G., Morin, J., Jacquin, F., Loridon, K., Quillet, M.C., Petit, A., Rameau, C., Lejeune-
Henaut, I., Huguet, T. and Burstin, J. 2006. Theor. Appl. Genet. 111: 1022-1031.

3. Ausin, I., Alonso-Blanco, C. and Martinez-Zapater, J.M. 2005. Intl. J. Dev. Biol 49: 689-705.

4. Barber, H.N. 1959. Heredity 13: 33-60.

5. Benlloch, R., Berbel, A., Serrano-Mislata, A. and Madueno, F. 2007. Ann. Bot. 100: 659-676.

6. Berbel, A., Navarro, C., Ferrandiz, C., Canas, L.A., Madueno, F. and Beltran, J.P. 2001. Plant J. 25:

441-451.

7. Beveridge, C.A. and Murfet, I.C. 1996. Physiol. Plant. 96: 637-645.

8. Bradley, D., Ratcliffe, O., Vincent, C., Carpenter, R. and Coen, E. 1997. Science 275: 80-83.

9. Cockram, J., Jones, H., Leigh, F.J., O'Sullivan, D., Powell, W., Laurie, D.A. and Greenland, A.J.

2007. J. Exp. Bot. 58: 1231-1244.

10. Corbesier, L., Vincent, C., Jang, S., Fornara, F., Fan, Q., Searle, I., Giakountis, A., Farrona, S.,
Gissot, L., Turnbull, C. and Coupland, G. 2007. Science 316: 1030-1033.

11. Foucher, F., Morin, J., Courtiade, J., Cadioux, S., Ellis, N., Banfield, M.J. and Rameau, C. .2003.

Plant Cell 15: 2742-2754.

12. Frances, S., White, M.J., Edgerton, M.D., Jones, A.M., Elliott, R.C. and Thompson, W.F. 1992.
Plant Cell 4: 1519-1530.

Pisum Genetics

2007—Volume 39

Review

13. Gottschalk, W. 1979. Pisum Newslett. 11: 10.

14. Hayama, R. and Coupland, G. 2004. Plant Physiol 135: 677-684.

15. Hecht, V., Foucher, F., Ferrandiz, C., Macknight, R., Navarro, C., Morin, J., Vardy, M.E., Ellis, N.,

Beltran, J.P., Rameau, C. and Weller, J.L. 2005. Plant Physiol. 137: 1420-1434.

16. Hecht, V., Knowles, C.L., Vander Schoor, J.K., Liew, L.C., Jones, S.E., Lambert, M.J.M. and Weller,

J.L. 2007. Plant Physiol. 144: 648-661.

17. Hofer, J., Turner, L., Hellens, R., Ambrose, M., Matthews, P., Michael, A. and Ellis, N. 1997. Curr.

Biol. 7: 581-587.

18. Imaizumi, T. and Kay, S.A. 2006. Trends Plant Sci 11: 550-558.

19. Jack, T. 2004. Plant Cell 16: S1-17.

20. Kalo, P., Seres, A., Taylor, S.A., Jakab, J., Kevei, Z., Kereszt, A., Endre, G., Ellis, T.H.N. and Kiss,

G.B. 2004. Mol. Genet. Genomics 272: 235-246.

21. King, W.M. and Murfet, I.C. 1985. Ann. .Bot. 56: 835-846.

22. Koornneef, M., Hanhart, C.J. and van der Veen, J.H. 1991. Mol. Gen. Genet. 229: 57-66.

23. Lifschitz, E. and Eshed, Y. 2006. J. Exp. Bot. 57: 3405-3414.

24. McClung, C.R. 2006. Plant Cell 18: 792-803.

25. Mizoguchi, T., Wright, L., Fujiwara, S., Cremer, F., Lee, K., Onouchi, H., Mouradov, A., Fowler, S.,
Kamada, H., Putterill, J. and Coupland, G. 2005. Plant Cell 17: 2255-2270.

26. Murfet, I.C. 1971. Austral. J. of Biol. Sci. 24: 1089-1101.

27. Murfet, I.C. 1971. Heredity 27: 93-110.

28. Murfet, I.C. 1973. Heredity 31: 157-164.

29. Murfet, I.C. 1975. Heredity 35: 85-98.

30. Murfet, I.C. 1992. Flowering Newslett. 13: 10-20.

31. Murfet, I.C. and Reid, J.B. 1993. In Casey, R. and Davies, D.R. (eds.) Peas: Genetics, Molecular
Biology and Biotechnology, CAB International, Wallingford, pp 165-216.

32. Platten, J., Foo, E., Elliott, R., Hecht, V., Reid, J. and Weller, J. 2005. Plant Physiol. 139: 1472-
1482.

33. Putterill, J., Laurie, R. and Macknight, R. 2004. Bioessays 26: 363-373.

34. Reid, J.B. 1979. Ann. Bot. 44: 163-173.

35. Reid, J.B. and Murfet, I.C. 1979. J. Exp. Bot. 28: 811-819.

36. Reid, J.B. and Murfet, I.C. 1984. Ann. Bot. 53: 369-382.

37. Reid, J.B., Murfet, I.C., Singer, S.R., Weller, J.L. and Taylor, S.A. 1996. Seminars Cell Dev. Biol. 7:

455-463.

38. Sato, S., Nakamura, Y., Asamizu, E., Isobe, S. and Tabata, S. 2007. Plant Physiol. 144: 588-593.

39. Singer, S., Sollinger, J., Maki, S., Fishbach, J., Short, B., Reinke, C., Fick, J. and Cox, L. 1999. Bot.

Rev. 65: 385-410.

40. Sullivan, J. and Gray, J. 2000. Plant Cell 12: 1927-1937.

41. Sung, S. and Amasino, R.M. 2005. Annual Rev. Plant Biol. 56: 491-508.

42. Taylor, S.A. and Murfet, I.C. 1993. Pisum Genetics 25: 60-63.

43. Taylor, S.A. and Murfet, I.C. 1994. Flowering Newslett. 18: 39-43.

44. Taylor, S.A. and Murfet, I.C. 1996. Physiol. Plant. 97:, 719-723.

45. Taylor, S.A., Hofer, J. and Murfet, I.C. 2001. Plant Cell 13: 31-46.

46. Taylor, S.A., Hofer, J.M., Murfet, I.C., Sollinger, J.D., Singer, S.R., Knox, M.R. and Ellis, T.H.

2002. Plant Physiol 129: 1150-1159.

47. Taylor, S.A. and Murfet, I.C. 2003. Physiol. Plant. 117: 100-107.

48. Weeden, N.F., Kneen, B.E. and Murfet, I.C. 1988. Pisum Genetics 20: 49-51.

49. Weller, J.L. and Reid, J.B. 1993. Planta 189: 15-23.

50. Weller, J.L., Murfet, I.C. and Reid, J.B. 1997. Plant Physiol. 114: 1225-1236.

51. Weller, J.L., Reid, J.B., Taylor, S.A. and Murfet, I.C. 1997. Trends Plant Sci. 2: 412-418.