How to map using the RI lines

The Recombinant Inbred (RI) lines were generated from a cross between the Arabidopsis ecotypes Columbia and Landsberg erecta (Lister and Dean, 1993) (with Columbia as the male parent). 300 lines were generated to be used for mapping. A large number of markers have been mapped using 100 of these lines. Mapping a new marker using these 100 will give an accurate map position relative to >1000 other markers however a rough map position can be determined with a much smaller number of these lines (20 - 30, see below). For very fine mapping the additional 200 lines can be used but markers in the area of interest would then need to be scored on the 200 lines.

One can only directly map something on these lines if it is polymorphic between Columbia and Landsberg erecta and its segregation is therefore scorable in the RI lines. QuantitativeTrait Loci (QTL), ie. loci affecting quantitative traits, can also be mapped as they are segregating in the RI population, although these may not be obviously polymorphic in the parental ecotypes.

Examples of markers, loci or traits that have been mapped.

i) RFLP markers

For example, cosmid (Nam et al, 1989) or lambda (Chang et al, 1988) clones, T-DNA or transposable element flanking sequences, cDNAs, RAPDs, ARMS markers, CAPS markers, random bits of DNA etc... which show an RFLP between Columbia and Landsberg erecta.

Larger probes (ie. cosmid or lambda clones) clearly increase the chances of finding an RFLP, however if many restriction fragments hybridize to the probe scoring of the polymorphism may be difficult, in addition the bands which are polymorphic may be cross-hybridizing sequences that map to different positions in the genome. Sequences flanking T-DNA/transposon inserts or cDNAs (ie 0.1 - 3kb) may be too small to show RFLPs at a reasonable frequency so many enzymes may need to be tested.

Enzymes that have been most useful for finding RFLPs between Columbia and Landsberg erecta are BclI, BglII, BstEII, CfoI, DdeI, DraI, EcoRI, EcoRV, HindIII, SacI and XbaI. Other restriction enzymes which could be tested are the six-base cutters with AT-rich recognition sequences (as the Arabidopsis genome is AT-rich). Alternatively restriction enzymes with four-base recognition sequences can also be tested. Use of small fragments to isolate larger fragments (from cosmid or lambda libraries) should also help in finding an RFLP.

ii) Phenotypic differences

Two examples are:-
a) The erecta mutation.
Landsberg erecta carries the erecta mutation and therefore has a short and erect stature, Columbia has the wild-type gene and is therefore taller and less compact. This phenotype is segregating in the RI population and was scored in the RI lines, and the data compared to the segregation data of the 67 RFLP markers. This resulted in a map position of the ERECTA gene relative to the RFLP markers.

b) Resistance/susceptibility to a fungal pathogen.
The downy mildew Peronospora parasitica infects Arabidopsis. For the isolate No-Co, there is a differential resistance response, with Columbia showing the sensitive phenotype (sporulation, no hypersensitive response (HR)) and Landsberg erecta, showing the resistant phenotype (HR/cell necrosis). This trait segregated as a single locus with the Ler allele dominant for resistance. In order to map the locus (RPP5) the RI lines were infected with the No- Co isolate and the response, either resistant (L) or sensitive(C), scored (see below). This data was compared to the segregation data of the 67 RFLP markers enabling the map position of RPP5 to be determined (Parker et al, 1993).

iii) Biochemical differences

The length and structure of the side chains of several glucosinolates vary between Columbia and Landsberg erecta. These differences can be detected using gas chromatography. Glucosinolate profiles of the RI lines were carried out and using this data the map positions of genes affecting the glucosinolate pathway were determined (Magrath et al 1994).

iv) Quantitative Trait Loci

A wide range of quantitative traits are segregating in the RI lines, including stature, rosette size, leaf number and shape, and flowering time. These traits were not necessarily different in the parental ecotypes but result from the many new combinations of alleles in the Columbia and Landsberg erecta genomes that have been brought together in the RI lines. These measurements can be used in conjunction with the segregation data for the RFLP markers to position the multiple loci influencing these traits (C. Lister and C. Dean, unpublished results; D. Marshall and M. Kearsey, unpublished results).

NOTE: concerning markers g4715-a and g4715-b:
The g4715 cosmid probe detects two mappable polymorphisms, only one of which was reported earlier as "g4715" on chromosome I. However, the Dean/Lister data includes both polymorphisms. "g4715-a" is identical to the original "g4715" marker on I. "g4715-b", the newer marker, is on chromosome V. The image file displaying the g4715 polymorphism shows "g4715-a".

Steps to Mapping a Locus

1. Identify an RFLP

Identify an RFLP or other polymorphic feature in the parental ecotypes (see above) or a quantitative trait which can be scored in the RI lines.

2. Sow out the RI lines.

Plants to be used for DNA preps can be grown either in the glasshouse and tissue culture (for leaves) or in liquid media in flasks (for mainly root material). Plants that are grown for phenotypic or biochemical examination may require special growing conditions or treatments to reveal/accentuate the differences between Columbia and Landsberg erecta (ie inoculation with fungus, different light regimes, etc..).

If only using 20-30 RI lines for mapping the following lines have been selected as having the highest frequency of recombination over the five chromosomes and therefore should be the most informative for mapping purposes:-

20 lines +10 lines (= 30 total)
CS/N1911 (33) 
CS/N1929 (115) 
CS/N1945 (190) 
CS/N1946 (191) 
CS/N1948 (194)  
CS/N1951 (217) 
CS/N1953 (231) 
CS/N1954 (232) 
CS/N1957 (238) 
CS/N1960 (245)
CS/N1963 (263) 
CS/N1966 (267) 
CS/N1968 (283) 
CS/N1969 (284) 
CS/N1970 (288) 
CS/N1971 (295) 
CS/N1974 (302) 
CS/N1978 (332) 
CS/N1984 (356) 
CS/N1989 (370) 
CS/N1900 (4) 
CS/N1901 (5) 
CS/N1903 (13) 
CS/N1913 (35) 
CS/N1915 (37) 
CS/N1959 (242) 
CS/N1975 (303) 
CS/N1985 (358) 
CS/N1988 (367) 
CS/N1990 (377)

3. Score the lines for your locus

i) RFLP markers. Harvest plant material (leaves or roots), make DNA. Do restriction digests, Southern blots, and hybridization experiments or do PCR reactions and run gels. Score (see below).

ii) Phenotypic difference. Score differential phenotype in the lines.

iii) Biochemical difference. Carry out the biochemical assay and score.

(iv) QTL Count or measure the trait under examination, usually recorded as a mean from 5-10 individuals/RI line.)

In all the above experiments do not forget to include the Columbia and Landsberg erecta parents at the same time, as controls.

4. Submit the data to NASC.

Please note that this service is no longer available due to a rapid decrease in requests following the publication of the sequence. We hope this does not cause any problems and thanks to all of the users that populated this map in the past

The text below is only recorded for historical reasons

RI mapping is conducted using Mapmaker (Lander et al 1987) and so the data has to conform to the Mapmaker format. We have designed an RI data submission form for data submission to help you send in the data in the correct format. We recommend that you use this form for data submission and that you fill in the information for each line. It has been designed to allow us to automatically enter your data and so circumvents any potential problems with typing errors.

However, we can also receive data by email Label your message RIDATA. If sending your information by email please adhere strictly to the following format, indicating the numbers of the RI lines used and including the data in (RI) numerical order.

Note: concerning plant order:
The 101 plants in the Dean/Lister population are presented in the following order. Each symbol corresponds to one an RI line (for example, "4" corresponds to CL4, which is Nottingham strain N1900.

PLEASE look at the RI data submission pages to correlate these with stock numbers. - they are NOT simply consecutive

4      5      13     14     17     19A    19B    25     29     30
32     33     34     35     36     37     46     52     53     54
59     62     67     68     71     79     84     90     107    113
115    123    125    131    160    161    166    167    173    175
177    179    180    181    182    188    190    191    193    194
199    209    214    217    231    232    235    237    238    240
242    245    253    257    259    263    264    266    267    279
283    284    288    295    296    297    302    303    311    321
332    342    345    349    350    351    356    358    359    363
367    370    377    378    386    390    394    395    397    398
Enter your scores for the RI lines data into a Text file (eg. MacExcel, or another text file) as follows:-
A = like Columbia parent
B = like Landsberg parent
H = heterozygote (if possible to score)
- = unscorable or no score

Marker names should start with a * followed by a letter (small or capital) ie. *w23 and should have a single space or tab between each entry. Names MUST BE LESS than EIGHT characters. Mapmaker, the programme we use for mapping, is not case sensitive and so reads "a" and "A" as the same thing. Please check the current names used for Loci, so as not to generate conflicts.

Your file should look like this:

RI line Number		4	5	13	14	17	25	29
*A1 A A B - A B B
*b2 B B B A H - B
*C4 B A - A B B A
Save as TEXT ONLY.

5. Enter data into Mapmaker.

If the data is sent to NASC for mapping then this section is for interest only. Mapmaker for the Macintosh and UNIX are somewhat different to run (Lander et al 1987). If the database of 67 markers is being used with these programs we strongly suggest that you READ THE INSTRUCTION MANUALS and DO THE TUTORIALS before attempting to run the program.

Use of the Kosambi mapping function (as opposed to Haldane) appears to give the best fit for recombination data generated in Arabidopsis (Koorneef and Stam, 1992). Kosambi mapping function takes into account the effects of interference; which means that after one recombination event has occurred it is less likely that a second one will occur in adjacent regions, in the same generation.

Before entering data into Mapmaker one needs to add two lines of information above the markers scores. The top line indicates what type of mapping population the data is from, the options depend on the version of Mapmaker being used (see below). The second line indicates the numbers of individuals, the number of markers scored and defines the genotype symbols.

If Mapmaker V1.0 is being used the data is considered as coming from an F2 segregating population. This requires using a LOD value of 6.0 (default is 3.0) for the group and three-point commands, to take account of the multiple rounds of recombination. The top line of the data file should read:-

data type f2 intercross

If Mapmaker V2.0 is being used on a UNIX one can run the data as coming from an RI population ( and therefore use the default LOD of 3.0). The default genotypes for this version are A=Columbia and B=Landsberg. The top line of the file should read:-

data type ri self

The second line indicates the number of individuals in the population (ie. 100), the number of markers being scored (ie. 46) and defines the genotype symbols used in the mapping data (ie. CLXYHU or AB-). The respective second lines will look something like:-

(V1.0) 100 46 0 0 CLXYHU (V2.0) 100 46 0 AB-

(PLEASE CHECK the manual of the version of Mapmaker being used that these lines are correct for that version, they do vary).

At the end the program will produce a list of the markers with the centimorgan distances and recombination fractions between them. On the Macintosh this can be converted into a map; this is also possible on the UNIX but the map produced is not so informative. If Mapmaker V1.0 has been used the the centimorgan distances will have to be recalculated, they are approximately two-fold too big as they were calculated for an F2 population. The formula is:-

r = R/2(1-R)
	where R is the recombination fraction.
	eg. the distance between two markers is 16.5cM (R=15.9%)
r = 0.159/2(1-0.159)
  = 0.094
  = 9.4cM


Chang C., Bowman J.L., DeJohn A.W., Lander E.S. and Meyerwitz E.M. (1988). Restriction fragment length polymorphism linkage map for Arabidopsis thaliana. PNAS 85 6856-6860

Koornneef M. and Stam P. (1992). Genetic analysis. In "Methods in Arabidopsis Research" (Koncz C.,Chua N-H and Schell J., eds.) World Scientific Publishing Co. Pte. Ltd., Singapore. 83-99

Lander E.S., Green P., Abrahamson J., Barlow A., Day M.J., Lincoln S.E., and Newberg L. (1987). Mapmaker: an interactive computer package for constructing primary genetic linkage maps of experimental and natural populations. Genetics 121 174-181

Lister C. and Dean C. (1993). Recombinant inbred lines for mapping RFLP and phenotypic markers in Arabidopsis thaliana. Plant Journal 4 745-750

Magrath R., Bano F., Morgner M., Parkin I., Sharpe A., Lister C., Dean C., Turner J., Lydiate D. and Mithen R. (1994) Genetics of aliphatic glucosinolates. 1. side chain elongation in Brassica napus and Arabidopsis thaliana. Heredity (in press).

Nam H-G, Giraudat J., den Boer B., Moonan F., Loos W.D.B., Hauge B.M. and Goodman H.M. (1989) Restriction fragment length polymorphism linkage map of Arabidopsis thaliana. Plant Cell 1 699-705

Parker J., Szabo V.,Staskawicz B., Lister C., Dean C., Daniels M. and Jones J. (1993) Phenotypic characterization and molecular mapping of the Arabidopsis thaliana locus RPP5 determining resistance to Peronospora parasitica. Plant Journal 4 821-831