PNL Volume 15 1983 RESEARCH REPORTS 43
A THREE-POINT LINKAGE ANALYSIS INVOLVING Am-1 — Af--I
Marx, G. A. NYS Agricultural Experiment Station, Geneva, NY, USA
This article adds to the unfolding account of the am-1 and am-2
genes and their associated seed disorders (6,7,8). With respect to
flower and axil color, both genes exert seemingly an identical effect.
Moreover, plants that are homozygous recessive for both am-1 and am-2
cannot be distinguished, on the basis of flower phenotype, from plants
that carry either gene alone. However, the two genes may be distin-
guished one from another on the basis of the seed disorder with which
each is peculiarly associated. They may also be distinguished by their
linkage relations, although thus far only am-1 has been mapped. The
linkage relations of am-1 are considered here.
All four of the loci relevant to this discussion, viz. A, Af, I,
Am-1 . are known to reside in chromosome 1, but _ is widely separated
from the other three. Am-1, discovered by de Haan (1), was found by
Wellensiek (10) to be linked with I at a map distance of about 26 units.
Lamprecht (4) confirmed the linkage but found widely varying crossover
values, ranging from a low of 9% to a high of 48%. Still, the mean of
all his crosses (28*) was close to that reported by Wellensiek. Linkage
values between.Af and X were reported by Khangildin (3), Snoad (9), and
Kielpinski (2) to be 45, 14, and 8 percent, respectively. Finally, Marx
(5) reported a relatively tight linkage between _f and Am-1 (7-8%).
Together, the foregoing two-point tests suggest the following order and
approximate map distances:
Am-1 8 Af 10 I
Apparently this is the first report of a three-point cross involv-
ing the above loci. Reciprocal crosses were made between two lines with
the following genotypes:
P1 a af i Am-1
P2 A Af I am-1 (sd-1)
F2 plants, 250 in all, were grown and scored in the greenhouse.
Additional seeds from the same cross will be planted in the field in the
summer of 1983 , but the essence of the experiment is already revealed in
the F2 data presented in Table 1. Moreover, there is usually less
chance of misclassifying the yellow (I/-) vs. green (i/i) cotyledon dif-
ference in greenhouse-grown plants than in field-grown plants where the
seeds often are subject to bleaching.
Two of the segregating traits in F2 could be discerned prior to
flowering, viz. foliar architecture (Af vs. af) and axil color. Axils
were either purple-violet (A Am-1). pink (A am-1), or colorless (a). No
af am-1 F2 recombinants were recovered because the two genes were linked
in repulsion and apparently there were too few plants to reveal cross-
overs. Since approximately one fourth of the population was recessive
for a. and since the Am-1 locus is hypostatic to a, segregation for
Am-1-am-1 could not he observed in a. plants, at least not directly.
At flowering, two classes of flower color could be distinguished:
Purple-violet (A Am-1) and white (A am-1, and a). These two phenotypes
were expected in an approximate ratio of 9 colored to 7 white. The ob-
served ratio of 125:125 yielded a Chi-square value of 3.97 (1 d.f.), a
value which is significant at 0.05 but not at 0.001.
PNL Volume 15 1983
All A. am-1 segregants exhibited the seed disorder and all such
plants were homozygous for I (I/I)None of the a I/I segregants ex-
hibited a seed disorder, yet, because of the close linkage between I and
am-1, all or nearly all of the a. I/I segregants would be expected to be
am-1. This, then, is a way to identify am-1 plants in an a. background
and shows that expression of the seed disorder is blocked in a. plants.
These results confirm progeny test results of A/a. am-1/am-1 sd-1 plants
presented earlier (8). Thus a is epistatic to am-1 not only with
respect to the flower color phenotype, but also with respect to the as-
sociated seed disorder. In contrast, b is hypostatic to am-1 and/or
am-2 with respect to flower color (i.e. am-1 b and am-2 b. flowers are
white) but epistatic to am-1 and/or am-2 with respect to their as-
sociated seed disorders. B-b_ did not segregate in the present crosses.
This was predictable on the basis that the seeds of the am-1 parent ex-
hibited the seed disorder and b. is thought to mask the expression of
seed disorder in am-1 plants just as it does in am-2 plants.
Because of the way this cross was constituted, the linkage rela-
tions among Am-1 --Af--I allow one to use the seed phenotype of one
generation to reliably indicate the plant genotype of the ensuing gener-
ation. Thus, i/i F3 seeds (either A or a) can be expected to produce af
F3 plants without the seed disorder. I/I seeds with A will produce Af
am-1 plants with the seed disorder and I/I seeds with a will produce Af
am-1 plants without the seed disorder. Heterozygous I/i (A/-) can be
used as a continuing source of segregating genotypes and as aid in
developing isogenic lines. The confidence that can be placed in these
predictions is, of course, a function of the intensity of the linkages.
In the present cross the probability of being correct is high because
the observed linkage intensity between Am and I was considerably
stronger than that reported previously. The af-i linkage gave an es-
timated recombination fraction of 3.7 + 1.2. Linkage chi-square for A
vs. Af and I was non-significant.
Since my original source of am-1 (L-3 from Dr. Lamm) exhibited fair
to poor seed set, I have selected for fully fertile, productive am-1
plants in order to be sure that the seed disorder is the direct,
pleiotropic effect of am-1 and not the result of some other cause.
Moreover, Lamprecht (4) suggested that am-1 plants were weaker than Am-1
plants, a suggestion I have not. been able to confirm. I now have fully
fertile, productive am-1 lines which, as was the case in the populations
discussed herein, give fully fertile F2 segregants in crosses.
The am-1 line used as a parent in the present crosses showed a
strong expression of the seed disorder in the greenhouse but when grown
in the field expression was considerably diminished. Upon return to the
greenhouse expression of this line was again strong, as was the expres-
sion in all the A am-1 F2 segregants derived from the cross with the
above line. Thus, the seea disorder phase of am-1 expression is in-
fluenced to some degree by the growth environment. What component(s) of
the environment that may operate to effect these differences in expres-
sion remains unknown. Soil pH is one possibility.
1. Haan, H. de 1930. Genetica 12:321-439.
2. Keilpinski, M. 1982. PNL 14:30.
3. Khangildin, W. V. 1966. Genetika (USSR) 6:88-96.
4. Lamprecht, H. 1954. Agri Hort. Genet. 12:38-49.
PNL Volume 15
RESEARCH REPORTS 45
Marx, G. A. 1969. PNL 1:9-10.
Marx, G. A. 1975. PNL 7:28-29.
Marx, G. A. 1978. PNL 10:38-40.
Marx, G. A. 1981. PNL 13:30-32.
9. Snoad, B. 1971. PNL 3:43.
10. Wellensiek, S. J. 1949. Genetica 24:71»-89.