Pisum Genetics
2006—Volume 38
Research Papers
Fs and U appear to be alleles of a locus near the end of linkage group V
Weeden, N.F.
Montana State Univ., Bozeman, MT, USA
Both Fs (violaceopunctata) and U (violet testa) are described loci that have been placed on the lower half of linkage group (LG) V as that linkage group is traditionally depicted (1, 8). Phenotypes produced by alleles at Fs and U have an interesting interrelationship because both produce anthocyanin pigmentation in the testa. Dominant alleles at Fs generally produce more or less intense spots of anthocyanin (usually 20 to 100) on the testa, whereas the U allele causes anthocyanin to be produced uniformly throughout the testa, resulting in a dark violet or ‘black’ seed. However,
occasionally in an Fs line seeds will be formed in which the spotted phenotype is replaced by a solid pigmentation that covers much or all the testa. This phenotype is referred to as ‘obscura’ (Fig. 1). The cause of the obscura phenotype is unclear because the next generation nearly always displays the spotted testa phenotype (1). Furthermore, a second allele of U, designated Ust, produces stripes or streaks of anthocyanin pigmentation on the testa, differing from a typical Fs pattern primarily by the streaks being larger and more
irregular than the spots produced by Fs alleles, as well as by their fewer number (0-5) on the testa.
Lamprecht (6) reported a recombination value between Fs and U of 23%, and this
Fig. 1 Different expressions of the ‘obscura’ phenotype (A, B, D) compared with a U phenotype (C). The seeds shown in E are from a single plant grown from a seed with an Fs testa phenotype. The obscura seeds in E are virtually indistinguishable from true-breeding ‘U’ seeds.
arrangement is reflected in Blixt’s classical
linkage map (1). However, in two other experiments he reported less than 1% recombination (4, 5) between the two phenotypes. Linkage analysis between Fs and U is complicated because the Fs phenotype is hypostatic to U and thus cannot be observed in seeds with the typical U phenotype. Furthermore, in lines homozygous Ust, not every seed displays the streaks characteristic of this allele. Thus, if relatively few seeds are collected from a plant possessing the Ust allele, it is possible to incorrectly score the plant as u/u. Finally, the obscura phenotype can complicate the scoring of phenotypes when both Fs and U are segregating in a population.
In order to determine the actual linkage intensity between Fs and U, I mapped both Fs and U relative to the isozyme locus Pgdc. Upon finding that both genes mapped to the same location within the error of the analysis, I attempted to reject the hypothesis that Fs, U and Ust are all alleles at the same locus, both by reviewing the literature and by performing the analyses described below. I am unable to reject the hypothesis and, indeed, developed evidence that neither U nor Ust are 100% penetrant when involved in crosses. Thus, the “recombination” observed between Fs and Ust by myself and other pea geneticists could merely reflect incomplete expression of the U or Ust phenotype. Although the history of the nomenclature for U and Fs is complicated, both genes being initially given different symbols [Ast for U (9) and P for Fs (3)], it appears U has precedence over Fs, and at this point it seems appropriate to retire the symbol Fs and identify the violaceopunctata phenotype in JI 261, A578-238 and many other lines as being produced by an allele of U, namely Ufs. This allele is dominant relative to absence of spotting (u) and codominant with Ust. A second locus (F)
Pisum Genetics                                                  2006—Volume 38                                           Research Papers
ostensibly able to produce the violaceopunctata phenotype, has been described previously (7,9) and has been assigned to LG III (6). The elimination of Fs should have no effect on the validity of F, although I have not been able to verify the existence of this latter gene.
Material and Methods
Several crosses, involving different sources of the alleles U, Ust or Fs were used in this analysis (Table 1). Each of the parents had been inbred several generations and consistently displayed the same testa phenotype in each generation. In two cases (WL1414 and WL1018), the line is homozygous a, and thus does not express Fs or U alleles. However, when crossed to a line that is homozygous A, the testa pattern can be observed in seed from the hybrid plant and in subsequent generations in seed of plants expressing the A allele.
Table 1. Populations used to examine the joint segregation of U, Fs and Pgdc
Cross designation
Parents of cross
Genotype of parents
Source of U
B77-257 x A578-238
fs, u x Fs, u
U not present
WL1238 x ‘Sparkle’
fs, Ust x [a] (fs, u)1
S. Blixt
WL1414 x WL808
[a] (Fs, u) x fs, U
P. s. ssp. abyssinicum
WL1018 x 87-19i-a
[a] (U) x Fs, u
S. Blixt
87-19I-a x WL1018
Fs, u x [a] (U)
S. Blixt
C98-54 x C98-1-12
U x Fs
N.F. Weeden
WL1238 x JI 261
fs, Ust x Fs, Ust
S. Blixt and P. s. ssp humile
Marx 15241 x Marx 15098
fs, oh x (Fs), U
G.A. Marx
C01-1b x A03-123
(Fs?), u/U x Fs, u
N.F. Weeden
1Parentheses indicate hidden phenotype of Fs, U or both that is not directly observable due to hypostatic interactions with other genes.
Plants were grown in the glasshouse under 16 hr daylength or for a relatively few populations, in the field at Bozeman, Montana, USA. Pods were allowed to dry on the plant to allow full development of testa pattern.
Results and Discussion
As testa pattern is determined by the maternal genotype, the seed produced from a cross should display the same phenotype as the maternal parent. Furthermore, seed produced on all F1 plants from either U x Fs or Fs x U crosses, should have the U phenotype. The former prediction was confirmed in all crosses. The phenotypes observed on seeds from the hybrid in each cross are presented in Table 2.
Table 2. Phenotype of seeds obtained from the hybrid plants in each of the crosses studied
Phenotype on seeds
Phenotypes observed
Phenotype of Parents
from hybrid
on seeds from F
fs, u x Fs, u
all seeds Fs
29 Fs, 10 fs
fs, Ust x a
all seeds Ust
53 Ust, 18 u, 29 a
a x fs, U
7 U, 3 Ust, 25 u, (Fs or fs)
0 U, 0 Ust, 9 Fs, 6 fs
a x Fs, u
all seeds U
4 U, 4 Fs, 7 a
Fs, u x a
U or U*1
37 U or U*, 10 Fs, 13 a
U x Fs
20 U and 14 Fs
4 U, 9 U*, 11 Fs
fs, Ust x Fs, Ust
all seeds Fs and Ust
Not tested
fs x U
3 U, 2 Fs
1 U, 1 U*, 2 fs, v. faint U
u/U x Fs, u
both seeds U*
U and Fs (backcross)
Asterisk indicates that the U phenotype only partially covered testa
Pisum Genetics
2006—Volume 38
Research Papers
Cross 1 (B77-257 x B578-238)
This cross was used primarily to confirm normal segregation of Fs and to determine recombination distance between Pgdc and Fs. The original two seeds produced from the cross both were fs, u, the same phenotype as the maternal parent. The hybrid plants generated from these two seeds produced only Fs seed (24 seed from one plant and 15 from the other). Of the 39 F2 plants grown, 29 produced only Fs seeds and 10 produced only fs seed. Joint segregation analysis of Fs and Pgdc gave a recombination distance between the two loci of 6 cM. Thus, this cross and F1 and F2 generations indicated that Fs was behaving as a single Mendelian factor, closely linked to Pgdc. Other crosses placed Fs distal to Pgdc relative to Acp1 (Weeden, unpublished).
Cross 2 (WL 1238 x ‘Sparkle’)
This cross was used primarily to confirm normal segregation of Ust and to determine recombination distance between Pgdc and Ust. Four seeds were produced from the original cross, and each had the maternal fs, Ust testa phenotype. The seed from the hybrid plant also displayed this same phenotype. If the white-flowered plants are excluded from consideration, the F2 population segregated approximately 3 Ust : 1 u (Table 2), and the calculated recombination distance between Pgdc and U was 7 cM. The locus Gp was also segregating in this population, and it mapped on the opposite side of Pgdc from U in agreement with (2). These results place U at approximately the same position as Fs, within the precision of the data from crosses 1 and 2.
Cross 3 (WL1414 x WL808)
This cross contained the violet testa (U allele) observed in many P. s. ssp. abyssinicum lines. The initial cross gave seeds lacking anthocyanin because the maternal parent was white-flowered. The F1 generation surprisingly produced mostly seeds lacking a solid violet testa, indeed most lacked any evidence of U expression with seven showing the solid violet pattern, three a partial violet testa, and 25 being either Fs or fs (a careful examination of the Fs phenotype was unfortunately not performed on this generation). When the F2 generation was grown, only 25 of the 35 lines produced seed and 10 of these were white-flowered, precluding the scoring of Fs or U. Of the colored-flowered plants producing seed, none were U or Ust, nine were Fs and six were fs. One of the Fs and two of the fs plants were from seeds with a U or partial U phenotype. The fs phenotype came from the P. s. ssp. abyssinicum parent, indicating that either all six of the fs F2 plants reflected recombination between Fs and U or that the U phenotype had been suppressed in these plants. The 9:6 ratio does not differ significantly from the expected 3:1 ratio (assuming lack of U expression), suggesting that segregation is not strongly distorted in this region of the genome.
Cross 4 (WL1018 x 87-19i-a)
The seed produced from the cross displayed the expected solid violet testa, as did the seed produced from the F1. Nearly half the F2 plants were homozygous for a, so that U and Fs could not be scored. Four of the F2 plants produced at least some seed with solid violet testa, whereas four F2 plants produced only Fs seed. However, two of the F2 plants scored as U produced some seed on which only part of the testa was solid violet, the remaining being violet spotted. Here again it appears that U is often incompletely dominant in segregating populations.
Cross 5 (87-19I-a x WL1018)
The hybrid seed possessed the Fs phenotype characteristic of the maternal parent. The seed from the F1 was a mixture of U and U plus Fs phenotypes. The U:Fs segregation ratio in the F2 did not differ significantly from the expected 3:1 ratio. However nine of the plants scored as U gave seed that only displayed partial fusion of the anthocyanin pigmentation into a solid pattern. Some of the seed collected
Pisum Genetics
2006—Volume 38
Research Papers
from these plants displayed only the spotted pattern typical of Fs genotypes. One plant gave three pods, one in which all seeds were U, one in which all seeds were Fs, and one in which the seeds were Fs but some had a partial fusion of the spots. These results mimic the spectrum of patterns observed in some Fs genotypes in which the obscura phenomenon is common. In addition, several of the plants scored Fs had slightly larger spots on the testa of some seeds, suggestive of a Ust pattern in combination with Fs. All of the F2 plants with an incomplete U pattern showed violet spots in the open regions. The lack of a double recessive phenotype can be explained by either WL1018 being Fs, U or with Fs and U being allelic.
Cross 6 (C98-54 x C98-1-12)
This cross used a line expressing the U phenotype (C98-54) that had produced solid violet seed coats for several generations, although the pedigree could not be traced back to either P. sativum ssp. abyssinicum or Lamprecht’s type line for U. The hybrid seed had the expected solid violet testa. However, the seed from the F1 was a mixture of phenotypes with some of the U seeds displaying only a partially violet testa and a significant proportion of the seeds exhibiting an Fs phenotype. Not all the F2 plants produced seed, but of those that did, 4 produced only seeds with solid violet testa, 9 produced seeds displaying various mixtures of partially solid violet testa and 11 produced only seeds with the Fs phenotype, which was sometimes very faint. In order to confirm the F2 seed testa phenotype, two seeds were taken from each F2 and grown to produce seed from the F3. In most cases the testa phenotype from F3 plants corroborated that from the F2. However, there were some notable exceptions. In two cases seed from the F2 that possessed an Fs phenotype gave seed with a U phenotype in the next generation. In another three cases, seed from the F2 that had barely discernable violet spots (although two had obscura markings) produced seed from the F3 that had clear Fs markings. In summary, this cross gave a slight but significant (c2 = 5.5, 1 d.f.) deficiency in the dominant (U) phenotype in the F2 even after correcting for the two lines that displayed U in the F3 but not in the F2 (Table 2), and there were a number of cases in which the expression of a dominant character (U or Fs) was observed in seed from F3 plants, when it had been lacking in the previous generation.
Cross 7 (WL 1238 x JI261)
This cross gave the expected Ust, fs phenotype on the one seed produced. The hybrid plant was semi sterile, and only four seeds were obtained. All these seeds displayed both the typical Ust streaks in a background of small violet spots, indicating that both Ust and Fs were expressed. I was not able to detect on the seed from the F1 generation the small streaks that characterize the Ust phenotype of JI 261. Rather all four seeds had the large blotches characteristic of WL 1238 together with many round spots attributable to Fs from JI 261.
Cross 8 (Marx 15241 x Marx 15098)
This cross was complicated slightly by the presence of recessive alleles at Oh (testa reddish-brown) and B (petals pink) segregating in the population. However, the critical finding for the purposes of this paper was that both seeds with the U phenotype and seeds with the Fs phenotype were produced from the hybrid plant. The known genotype of the hybrid was u/U, b/B, oh/Oh. I am uncertain whether Marx 15098 has violet spots obscured by the solid violet color of the testa. However, the seeds from the hybrid would all be expected to display the U phenotype. The presence of seeds with only violet spots further indicate that U is not 100% penetrant in some crosses. Seeds collected from two F2 plants derived from seeds with U phenotype and two F2 plants derived from seeds with Fs phenotype were, respectively, U, pale U (with no violet spots), fs with a faint obscura, and fs with a faint pinkish hue. All four F2 had wildtype (B) flowers. Finally, when a typical U phenotype seed from the first of the F2 plants mentioned was used to produce F3 seed, this seed was uniformly Fs. Thus, in this cross the U phenotype was lost by the third filial generation, appearing to transform into an Fs pattern despite the U allele having been
Pisum Genetics
2006—Volume 38
Research Papers
derived from a line that stably expressed the U phenotype for several generations. The loss of the Fs phenotype in some lines when going from the F2 to the F3 generation is also interesting, although it could be explained by simple segregation, or interactions with b or oh.
Cross 9 (C01-1b x A03-123)
C01-1b was an F1 plant produced from the cross B5 (U) x B77-257 (fs, u). The line B5 shared the same U allele as C98-54 in cross 6. The cross of this hybrid with A03-123 (Fs) gave two seeds both of which display a testa that was partially solid violet and partially violet spots. Selfed seed from C01-1b was typical U (completely solid violet). The two hybrid seeds were planted, and one produced all U seeds, whereas the other produced all Fs seed. This last results can be interpreted as being consistent with the segregation at U from C01-1b. However, because the original hybrid seeds were produced on a plant that was heterozygous at U, both of these seeds should have been uniformly solid violet. It again appears that U is not completely penetrant in some genetic backgrounds.
The results from the above crosses clearly demonstrate that the solid violet testa phenotype described for the U allele is not completely penetrant in many crosses. This behavior does not appear to be dependent on the source of the U allele, for lines from the Weibullsholm collection, the Marx genetic stocks collection, my own collection and the taxon P. s. ssp. abyssinicum all showed incomplete expression in at least one cross. At present, the specific genetic background (if any) that produces the incomplete expression is not obvious.
Joint segregation analysis between the locus Pgdc and either Fs or U revealed nearly identical recombination frequencies between the isozyme marker and each of the seed testa phenotypes, placing both Fs and U distal to Pgdc (in agreement with previous studies) and within 1 to 2 cM of each other. This result is in disagreement with the position of U on LG V on certain linkage maps for pea (1) but agrees well with some of Lamprecht’s data (3, 4). When the two markers are separated on LG V, U is always placed distal to Fs, and usually forms the most distal marker on that arm of the linkage group. I suggest that most recombinants identified between Fs and U in earlier studies are a result of incomplete expression of the U phenotype and are apparent recombinants rather than real. The only evidence for a significant recombination frequency between Fs and U is the appearance of the fs phenotype on seeds in the F2 of cross 3. Such seed could be interpreted as being produced from a plant that was recombinant between Fs and U on both homologous chromosomes. However, the F1 of cross 3 already displayed significant loss of U expression, suggesting that a general loss or suppression of U expression might be a more conservative explanation for the fs phenotype. Indeed, a study of a large population produced from backcrossing a P. sativum ssp abyssinicum x ‘Sparkle’ hybrid (U/u, fs/fs) to other white-flowered cultivars (u/u) revealed a complete absence of U phenotypes in the seed from 26 BC1F3 plants that displayed violet flowers (data not presented). The only anthocyanin markings on testa of seed from this generation were very faint obscura patterns in some lines. This result suggests that the solid violet testa in P. s. ssp. abyssinicum is not particularly stable in crosses and possibly can be transformed to obscura, a phenotype associated with the Fs locus.
As heterozygotes from all sources of U can show incomplete penetrance, the mapping of U and the use of U as an anchor marker become problematic. If U maps near Fs, an allelism test is called for, yet with Fs being hypostatic to U, with the capability of Fs converting to an obscura phenotype very similar to an incomplete expression of U, and with the expression of U being erratic in many crosses, it becomes difficult perform such a test. To my knowledge, the only lines used in the above experiments that are definitely fs, U are the violet testa P. s. ssp. abyssinicum accessions, and this taxon displays the greatest loss of penetrance of the U phenotype in seed from F1 plants. In all other crosses between a line with an Fs phenotype and a line with a U phenotype I have yet to isolate a derivative with an fs phenotype lacking
Pisum Genetics
2006—Volume 38
Research Papers
some type of obscura pattern. My review of the literature also fails to identify an allelism test in which the problems of incomplete penetrance and the obscura phenomenon have been clearly overcome.
It has always been the responsibility of the investigator, when describing a new gene, to perform the necessary allelism tests. In the case of Fs and U, both symbols are already accepted in the literature despite an apparent absence of a rigorous allelism tests, due to the complications described above. Again because of these complications, I have been unable to provide clear evidence that the two phenotypes represent alleles at the same locus. However, given the facts that (1) the phenotypic expression of both Fs and U is unstable and produces similar ‘off types’ and (2) in independent crosses phenotypic segregation places Fs and U within the same 1-3 cM region on LG V, the most conservative alternative at present is to treat Fs and U as the same locus. Hence, I recommend that Fs be referred to as Ufs until a clear demonstration is available that it represents a distinct locus.
1.  Blixt, S. 1972. Agri Hort. Genet. 30: 1-293.
2.  Irzykowska, L, Wolko, B. and Swif cicki, W.K. 2001. Pisum Genetics 33: 13-18.
3.  Kajanus, B. and Berg, S.O.. 1919. Ark. f. Botanik 15: 1-18.
4. Lamprecht, H. 1942a. Hereditas 28: 143-156.
5.  Lamprecht, H. 1942b. Hereditas 28: 157-164.
6.  Lamprecht, H. 1961. Agri Hort. Genet. 19: 360-401.
7.  Tschermak, E. von. 1912. Z. ind. Abst. u. Vererbungsl. 7: 81-234.
8.  Weeden, N.F., Ellis, T.H.N., Timmerman-Vaughan, G.M., Swifcicki, W.K., Rozov, S.M. and Berdnikov, V.A. 1998. Pisum Genetics 30: 1-4.
9. White, O.E. 1917. Proc. Amer. Phil Soc. 56: 487-588.