AFLPs have been used to assess genetic diversity in germplasm collections of sugarcane and close relatives [49–52], and to build linkage maps [19, 53]. In light of these studies, AFLPs have been informative in generating a substantial amount of unambiguous polymorphic markers. For example, Andru et al. reported 64 AFLP restriction enzyme-primer combinations, and detected 816 polymorphic loci; in the present study, 22 combinations revealed 685 polymorphic loci. We suggest the differences were the result of the genotyping population, the first with increased homozygosity (S1 progeny) relative to a segregating F1 population. In both studies, AFLPs were a viable marker to build a scaffold for other marker types, including expressed sequence-based markers, which could be positioned on the scaffold. Furthermore, this scaffold is particularly important in sugarcane, as expressed sequences are physically too far apart.
From the 22 enzyme-selective primer combinations selected in this study to detect AFLPs, 14 were previously tested to build maps for the ‘R570’ cultivar , the S. officinarum ‘IJ 76-514’ clone, and ‘Q165’ , ‘Q117’ and ‘MQ77-340’  cultivars as well as for the F1 population ‘SP80-180’ x ‘SP80-4966’ . Based on shared AFLPs and SSRs, the map comparisons of ‘R570’, ‘Q165’, ‘Q117’, and ‘MQ77-340’ cultivars were done ; several co-segregation groups (CGs) were aligned, and homo(eo)logous groups (HGs) associated, which received the same designation. As several common alleles were positioned, therefore suitable data should be available to construct a reference map for sugarcane commercial varieties, despite the pedigree complexity. For instance, a comparison between the map built based on the present study and the one published by Pastina et al. is possible. Both are integrated linkage maps that share some common SSRs. In both maps, ESTA53, ESTB94, ESTB100 and ESTB118 were assigned to HGI as well as ESTA48 was mapped in the same grouping, HGIII. In addition, HGV contains three CGs that share the ESTB60 locus, which was assigned to HGVII of Pastina’s map . This suggests a possible correspondence between these HGs.
Several authors have applied SSR markers to estimate linkage in sugarcane [19, 30, 32, 33]. Due to the multiallelic nature and relative abundance of SSRs in plant genomes, they have utility to identify HGs in polyploid species . Based on this principle, Rossi et al. identified 66 CGs (of 128) assembled into seven HGs from the French cultivar ‘R570’ linkage map. Similarly, Aitken et al. identified 136 CGs assembled into eight HGs from the Australian cultivar ‘Q165’ linkage map. Oliveira et al. identified 120 CGs (of 192) assembled into 14 HGs from a map of the progeny derived from a single cross between ‘SP80-180’ and ‘SP80-4966’.
Here, we compiled 41 CGs (of 92) into seven putative HGs. Interestingly, six EST-SSR loci were duplicated within chromosomes CV22, CV38, CV78, CV100, ESTB14, and ESTB94, which were positioned on HG I and IV (Figure 1). Duplicated genomic regions were reported to occur in various sugarcane maps [23, 28, 32, 33], and are possibly a consequence of the multispecific origins of the modern cultivars. Structural genomic rearrangements, including the movement of transposable elements (TEs) may also explain the duplications .
We used an innovative approach by mapping transposon-based markers in sugarcane using the NBS-profiling technique. In other plant species, TEs have been mapped using SSAP. This marker system was applied in barley (namely the TE BARE-1) , wheat (TEs, Wis2A-1 A and BARE-1) , lettuce, (Tls1 and Tls2) , and tomato (ToRTL1
T265 and Tnt1) . Due to the advanced knowledge in tomato genetics, it was possible to determine that polymorphic insertions were primarily located in the centromeric regions. Both the above-mentioned approaches increase the available information regarding retrotransposon distribution over plant linkage groups. The NBS-profiling protocol efficiently targeted scIvana_1 retrotransposon sequences and, at the same time, produced a polymorphic multilocus marker profile that was enriched for these sequences. Both the SSAP approach and the NBS-profiling technique investigate polymorphic restriction sites and the presence or absence of the retrotransposon sequence. SSAP uses two restriction enzymes (one with frequent cut sites and the other with rare cut sites), generating a large number of DNA fragments before performing the selective amplification. PCR results from the use of a primer that is complementary to the adaptor sequence and other complementary to the retrotransposon sequence. Since we have to choose two enzymes which have no recognition sites for restriction in the retrotransposon sequence, it reduces the number of combinations (enzyme/primer) to be tested. The NBS-profiling technique only uses one restriction enzyme (with frequent or rare cut sites), generating a small number of DNA fragments to be selected and consequently be stained and visualized separately in the gel. This is especially important considering the large genome size of sugarcane. Using enzymes that have no recognition sequences in the retrotransposon scIvana_1 (combined with primers complementary to its sequence), it was possible to estimate the number of copies of this element in the genome.
Among the 16 restriction enzyme-primer combinations used to amplify scIvana_1, the enzymes DraI and SspI resulted in an increased number of scorable bands and polymorphic loci compared to AluI and RsaI. Earlier studies have shown rare-cutting enzymes such as DraI and SspI are more suitable for restricting sugarcane DNA due to its large size. However, enzymes that frequently cut sugarcane DNA have the potential to generate an enormous number of fragments, and consequently affect other protocol steps, and subsequent results. All enzyme-primer combinations resulted in non-amplification for scAle_1. Primer design was challenging due to sequence diversity among scAle_1 copies. When amplicons were obtained, all combinations resulted in ∼ 112 bands per gel, which prevented polymorphism identification. Alternatively, a reduction in the number of bands per gel can be reached by adding selective bases at the 3'-end of PCR primers, as previously demonstrated in barley . Queen et al. used SSAP to study the elements Wis2A-1A and BARE-1 in wheat, and four selective nucleotides were added as an attempt to reduce the number of amplified fragments. Both elements are known to have 1,000 copies in the wheat genome, and good results were obtained for genotyping and mapping when applying this strategy. Besides this, we should try to produce markers derived from scAle_1 subfamilies, therefore having an estimate of the number of copies of each subfamily in the sugarcane genome.
The mean number of amplicons obtained by restriction enzyme-primer combinations was very similar to the number estimated by molecular methods; 40 to 50 copies of scIvana_1 were detected in the sugarcane genome, and scAle_1 exceeded 1,000 copies . These results were congruent with 56 scIvana_1-based loci positioned on the linkage map, exhibiting preferential cluster distribution along 21 CGs. In large-genome cereals, Bennetzen  reported retrotransposon distribution as nested insertions in highly heterochromatic transposon clusters. Later, authors reported there appears to be some clustering of TE BARE-1/Wis-2-1 A-based markers on the linkage map of wheat , and transposon cluster interference with recombination machinery acting in adjacent euchromatic regions in maize . When additional genomic data is available for sugarcane, it will be interesting to investigate if genes adjacent to retrotransposon clusters are less recombinogenic. Dooner and He  suggested the more condensed chromatin state of retrotransposon clusters in maize might interfere with recombination machinery access in adjacent euchromatic regions. Additionally, in a recent review published by Kalendar et al. authors indicated that at least in cereals and citrus retrotransposons are often locally nested one into another and in extensive domains that have been referred to as ‘retrotransposon seas’ surrounding gene islands.
We are possibly facing an association between clustered retrotransposon sequences, the inhibition of DNA recombination, an explanation of the small map distance between adjacent retrotransposon-based markers, and the element copy number in plant the genome. This should explain scIvana_1 properties, such as low copy numbers (~60) with expression and mobility under strict control, conversely against the properties of scAle_1 retrotransposons. Therefore, mapping scAle_1 element is of great interest, as well as the location of these two elements in chromosome regions.
The segregation results presented here independently indicated that AFLPs, EST-SSRs, or scIvana_1-based loci were consistent with the outcome of former studies [26, 28, 32, 53, 62] i.e. most markers (~70%) were SDMs. Furthermore, a substantial number were unassigned markers, in addition to variation in the marker number per CG.
The genetic map constructed here (‘IAC66-6’ x ‘TUC71-7’) has 546 SDMs covering 4,843 cM that were ordered in 92 CGs, with a marker density of 8.87 cM. The genetic map recently published by Pastina et al. (‘SP80-180’ x ‘SP80-4966’) has 317 markers covering 2,468 cM that were ordered in 96 CGs, with a marker density of 7.5 cM . These are both integrated maps constructed based on segregating F1 populations. Note that the number of CGs is somewhat high in Pastina’s map, but it is shorter and denser. Cultivar maps are established using self-fertilized populations, and therefore are not comparable to other maps built based on different backgrounds. For instance, the ‘LCP 85-384’ map has 784 markers covering 5,617 cM that were assigned to 108 CGs, with a marker density of 7.16 cM , and ‘Q165’ map has 910 markers covering 9,058 cM that were assigned to 116 CGs, with a marker density of 9.95 cM . Note that, in this case, the shorter and denser map is the one with a low number of CGs.
Enhancement of sugarcane genetic maps should include additional segregation ratios in mapping analyses, and an increased number of informative SNP- and SSR-loci segregating in larger populations. In addition, there is a need for meiotic studies that it is an important component of future studies in deciphering the genetic configuration of sugarcane genotypes.
Finally, it is important to note that the parents of the mapping population differ in response to the Sporisorium scitamineum infection; therefore we expect that offspring segregate for this trait. Consequently, the genetic map established here should be used to localize quantitative loci. It will certainly contribute to a better view on the genetic architecture of smut resistance in sugarcane, as little is known on this subject [63, 64]. As recently shown in Pastina et al. integrated genetic maps are useful for mapping QTLs. Based on interval mapping and mixed models, authors map QTL effects on a segregating progeny from a cross between two pre-commercial cultivars. The same approach should be interesting to be applied using the present map that includes retrotransposon-based markers. Moreover, we should improve McNeil et al.’s strategy  by aligning marker sequences tightly linked to QTLs for smut resistance with data from the sugarcane genome sequencing project currently underway .