Ac-st and automutagenesis of Kn1-2F11
--Erik Vollbrecht and Sarah Hake

Knotted1 (Kn1) encodes a homeodomain-containing protein that is expressed in vegetative and floral meristems. Dominant mutations (Kn1) that alter leaf development apparently still provide Kn1 function: Kn1 mutants produce wildtype gene product (KN1) and express KN1 in meristems in a wildtype pattern. Additional, ectopic expression of KN1 in leaves causes the knotted phenotype in the mutants. A small, x-ray generated deletion of chromosome 1 that includes Kn1 (kn1-del) is lethal to the pollen but transmits through the female, and is an embryonic lethal when uncovered by TB-1La. Although kn1-del does not delete RFLPs closely linked to Kn1, the extent of the deletion is unknown and it may include additional loci. Thus while analysis of the wildtype and dominant mutant alleles has revealed much about the role of Kn1 in development, no loss of function mutation at the locus has yet been recovered.

The Kn1-2F11 mutation was used in a mutagenesis experiment designed to generate germinal derivatives with loss of Kn1 function. Kn1-2F11 is caused by a Ds2 insertion in an intron 150bp from an intron-exon junction. Transposon induced mutant alleles serve as excellent starting material for mutagenesis experiments. The resident transposon can act as a mutagen to create DNA alterations through events such as imprecise excisions, intragenic transpositions or rearrangements. One can screen for changes from the mutant phenotype, such as reversion to wildtype or alteration to a novel mutant phenotype. Critical to this approach is a mutant phenotype that is easily scored and reliably manifests when the mutation is present. Kn1-2F11 confers a mild knotted phenotype that is evident principally in the first one to three seedling leaves. The Kn1-2F11 mutation is also Ac dependent: penetrance and expressivity of the mutation increase markedly when an active Ac element is present in the genome (Vollbrecht and Hake, MNL64:4). Interactions between transposase and Ds2, and not transposition-associated DNA changes, appear to underlie the Ac dependence of Kn1-2F11 expressivity. In our bz2-m background, penetrance of Kn1-2F11 is 0-0.1 when no Ac is present (0% to 10% of those plants containing Kn1-2F11 manifest the knotted phenotype). When a standard Ac is in trans, penetrance increases to between 0.4 and 0.8. Most standard Ac elements are similar in their effect on Kn1-2F11 when compared in similar genetic backgrounds. We recently discovered that McClintock's stabilized Ac (Ac-st) is exceptional in its effect on Kn1-2F11. When Ac-st is present, penetrance of Kn1-2F11 is nearly 100%, and expressivity increases dramatically such that both seedling and adult leaves display a severe knotted phenotype. When Ac-st is present, Kn1-2F11 satisfies the criteria for an automutagenesis of the type described above. The data reported here are results of a pilot mutagenesis experiment, carried out in the field in which the Ac-st effect was first observed.

Given that kn1-del does not transmit through the male, this experiment used the Kn1-2F11 parent primarily as the female, although Kn1-2F11 was also used as the male in a few crosses. The basic cross was:
    F0 (Ac11/-; Kn1-2F11; bz2-m; R) X (kn1; bz2-m; Ac-st/-; r-m3/R)
    F1 All (Kn1-2F11/kn1; bz2-m); ±Ac-st and ±Ac11.

Ac11 is a standard Ac element present in the bz2-m tester stock, and causes early transposition in the bz2-m background (Dawe and Freeling, MNL65:33). A few Kn1-2F11 lines containing other standard Ac elements were in the field and were also used. Standard Ac is included in the Kn1-2F11 parent to induce transposition of the resident Ds2 and thus potentially generate change of state derivatives, and Ac-st is in the outcross parent to enhance the knotted phenotype when Kn1-2F11 transmits unchanged to the F1. F1 kernels showing aleurone mutability (bronze to purple) were sorted into two classes: (1) Ac-st kernels: those exhibiting a spotting pattern typical of Ac-st (± a superposed Ac spotting pattern; Ac-st and Ac show little or no dosage interaction and can be scored simultaneously) and (2) Generic mutable kernels: those exhibiting a spotting pattern that was either typical of standard Ac dosage or unclassifiable. Kernels were planted in the greenhouse and seedlings were screened for alterations from the Kn1- 2F11 phenotype, i.e., for wildtype or novel individuals. In most families the Ac-st class showed sufficiently high Kn1-2F11 penetrance to suggest non-knotted individuals had truly undergone changes at the kn1 locus (Table 1). Penetrance in the "generic mutable" class was lower and only a few of these families were useful (Table 2). In reciprocal crosses involving the Ac11 lines, the recovery of wildtype was lower through the male (~4.1%, Table 3) than through the female (7.5%, Tables 1 and 2).

Wildtype seedlings were transplanted, tissue samples were taken for Southern analysis of genomic DNA and plants were eventually selfed. Hybridization of genomic DNA blots with a series of Kn1 fragments scanned 11kb of DNA. This region includes most of the Kn1 transcription unit, and 1.5kb of 5' sequences. Southern analysis identified 2 self contaminants (showed no outcross parent RFLPs), 23 non-excision individuals that still contained the Kn1-2F11 mutation, and 75 excision individuals that no longer contained Ds2 (Table 4). The 75 excision plants showed no novel RFLPs within the 11kb scanned by the genomic Southerns. No intragenic transpositions, deletions or rearrangements were detected. Southerns would have detected even short-range transpositions, as the restriction enzyme used (BamHI) cuts within Ds2. To date, all 75 excision events appear to be reversions to kn1 that transmit normally and confer a wildtype phenotype. Molecularly corrected reversion frequencies per chromosome still proved higher through the female than through the male (5.3% vs. 3.3%, Table 4). The higher frequency of reversions conditioned by Ac11 (5.3%) compared to other Ac elements (2.3 % combined) could reflect Ac11-induced somatic ear (or tassel) sectors due to early transposition events, or could reflect an increase in Ac11-induced transpositions. Our data do not favor one interpretation over the other.

No germinal derivatives with imprecise excisions, rearrangements or intragenic transpositions were detected in this experiment. We assume that some DNA rearrangements or alterations would cause loss of Kn1 function and would be recessive to wildtype kn1. Lack of recovery of a recessive was not likely due to inability of kn1 mutations to transmit, since kn1-del transmits normally through the female. While our Southern analysis may not have detected imprecise excisions changing ?50bp (the excision BamHI fragment is 5kb), if they do exist in the excision population we can only infer that either such alterations (within an intron) do not affect Kn1 gene expression or their effects are unimportant for screened aspects of the phenotype. We similarly assume that some intragenic Ds2 transpositions would result in kn1 (or kn1-m) alleles. Of the many (>12) transposon-induced Kn1 dominants that have been recovered, all but one have insertions into the same intron, near the site of the Ds2 insertion in Kn1-2F11. Thus we infer the intron region is unique, perhaps containing sequences relevant for Kn1 regulation, and insertions elsewhere within the gene would likely not cause the knotted phenotype and perhaps cause loss of Kn1 function. We may have missed intragenic transpositions in our screen if they conferred a knotted phenotype. It was reported for Wx-m5 that intragenic Ds transpositions (unidirectional, within a ~4kb region) occurred at a frequency of 5.5 x 10-4 (Weil et al., Genetics 130:175-185, 1992). Although our experiment would have detected bidirectional Ds2 transpositions within an 11kb interval, our population size may have been too small to generate such derivatives in this pilot experiment. The results of this experiment, however, suggest the Kn1-2F11 + Ac-st system can be utilized to generate and recover germinal change of state derivatives, including the elusive kn1 mutation. We have begun a much larger scale reversion experiment, with Kn1-2F11 plants containing one dose of Ac11 crossed by homozygous Ac-st lines. 15,000 F1 plants were screened in the field and roughly 500 putative wildtypes identified. Wildtype plants were selfed and F2 progeny screening is in progress.

Table 1. F1 seedling phenotypes, kernels with Ac-st spot pattern.
 

Kn1-2F11 transmitted as female.
Ac element^	family #	# kernels	# wt sdlgs	% wt
Ac11	E29	62	5	8
	E30	75	9	12
	E31	50	3	6
	E32	60	3	5
	E33	90	8	9
	E34	40	4	10
	E35	35	2	6
	E36	18	3	17
	E59	30	4	13
	Total Ac11	460	41	8.9
wx-m9	E45	40	3	7.5
	E46	24	5	21*
	E47	80	14	18*
wx-m7	E48	14	5	36*
	E51	76	6	7.9
r-nj:m	E53	54	many	high*
	E54	58	many	high*
	E55	55	many	high*
	E56	67	14	21*
	E57	38	2	5
	E58	27	0	0
	Total rnj:m	65	2	3.1
bz1-m2	E49	80	1	1.3
Total, as female**		721	53	7.4
Kn1-2F11 transmitted as male.
Ac11	E37	64	4	6
	E38	24	2	8
	E39	40	2	5
	E40	34	1	3
	E41	39	1	3
	E42	23	1	4
	E43	40	7	18*
	E44	24	0	0
Total, as male**		248	11	4.4

Table 2. F1 seedling phenotypes, generic mutable kernels.***
 

Kn1-2F11 transmitted as female.
Ac element^	family #	# kernels	# wt sdlgs	% wt
Ac11	E32	120	11	9
	E59	90	6	7
wx-m9	E46	90	7	8
r-nj:m	E57	42	3	7
Total, as female**		342	27	7.9
Kn1-2F11 transmitted as male.
Ac11	E37	185	4	2
	E38	135	8	6
Total, as male**		320	12	3.8

Table 3. Combined F1 phenotype totals.**
 

Kn1-2F11 as female	1063	80	7.5%
Kn1-2F11 as male	568	23	4%
Total, female + male	1631	103	6.3%

Table 4. Southern analysis of F1 plants (putative revertants).**
 

Summary of Southern data for 103 seedlings.
3	no DNA
2	self contaminants
23	Contained Kn1-2F11 (low expressivity)
75	revertants
103	total

Net reversion frequencies of Ds2 from Kn1-2F11 for different Ac's.

Ac element^		# kernels	# wt	% wt^^
Ac11	as female	670	47	7
	as male	568	19	3.3
	total Ac11	1238	66	5.3
wx-m9	as female	130	5	3.8
wx-m7	as female	76	1	1.3
r-nj:m	as female	107	2	1.9
bz1-m2	as female	80	1	1.3
Totals	as female	1063	56	5.3
	as male	568	19	3.3
Combined molecular total		1631	75	4.6

NOTES on tables:
^Refers to the Ac element effecting transposition in the F0.
^^Indicates reversion frequency per chromosome.
*When occurrence of WT seedlings was > 15%, we assumed poor
penetrance (<85%) rather than high reversion, and these families
were not investigated further.
**Families of assumed poor penetrance (see above) are not included.
***Generic mutable refers to spotted kernels in which unambiguous
classification of mutability pattern (Ac-st or Ac) was not possible.

Return to the MNL 67 On-Line Index
Return to the Maize Newsletter Index
Return to the MaizeGDB Homepage