Variations in gene expression have long been hypothesised to be the major cause of individual differences. An initial focus of this research thesis is to elucidate the genetic regulatory architecture of gene expression. Expression quantitative trait locus (eQTL) mapping analyses have been performed on expression levels of over 22,000 mRNAs from three tissues of a panel of recombinant inbred mice. These analyses are "single-locus" where "linkage" (i.e. significant correlation) between an expression trait and a putative eQTL is considered independently of other loci. Major conclusions from these analyses are: 1. Gene expression is mainly influenced by genetic (sequence) variations that act in trans rather than in cis; 2. Subsets of genes are controlled by master regulators that influence multiple genes; 3. Gene expression is a polygenic trait with multiple regulators. Single-locus mapping analyses are not designed for detecting multiple regulators of gene expression, and so observation of multiple-linkages (i.e. one expression trait mapped to multiple eQTLs) formed the basis of the second objective of this research project: to investigate the relationship between multiple-linkages and genotype pattern-association. A locus-pair is said to have associated genotype patterns if they have similar inheritance pattern across a panel of individuals, and these are attributed to one of fours sources: 1. linkage disequilibrium between loci located on the same chromosome; 2. non-syntenic association; 3. random association; 4. un-associated. To understand the validity of multiple-linkages observed in single-locus mapping studies, a newly developed method, bqtl.twolocus, is applied to confirm two-locus effects for a total of 898 out of 1,233 multiple-linkages identified from the three studies mentioned above as well as from seven publicly available eQTL-mapping studies. Combining these results with information of genotype pattern-association, a subset of 478 multiple-linkages has been deduced for which there is high confidence to be real.

Paget's disease of the bone is a skeletal disorder of unknown cause. This disease is characterised by excessive and abnormal bone remodelling brought about by increased bone resorption followed by disorganised bone formation. Increased bone turnover results in a disorganised mosaic of woven and lamellar bone at affected skeletal sites. This produces bone that is expanded in size, less compact, more vascular, and more susceptible to deformity or fracture than normal bone. Symptoms of Paget's disease may include bone pain, bone deformity, excessive warmth over bone from hypervascularity, secondary arthritis, and a variety of neurologic complications caused in most instances by compression of the neural tissues adjacent to pagetic bone. Genetic factors play a role in the pathogenesis of Paget's disease but the molecular basis remains largely unknown. The identification of the molecular basis of Paget's disease is fundamental for an understanding of the cause of the disease, for identifying subjects at risk at a preclinical stage, and for the development of more effective preventive and therapeutic strategies for the management of the condition. With this in mind, the aim of this project is to identify genetic loci, in a large pedigree, that may harbour genes responsible for Paget's disease of bone. A large Australian family with evidence of Paget's disease was recruited for these studies (Chapter 3). This pedigree has characterised over 250 individuals, with 49 informative individuals affected with Paget's disease of bone, 31 of whom are available for genotypic analysis. The pattern of disease in these individuals is polystotic, with sites of involvement including the spine, pelvis, skull and femur. Although the affected individuals have a severe early-onset form of the disease, the clinical features of the pedigree suggest that the affected family members have Paget's disease and not familial expansile osteolysis (a disease with some similarities to Paget's disease), as our patients have extensive skull and axial skeletal involvement. The disease is inherited as an autosomal dominant trait in the pedigree with high penetrance by the sixth decade. Due to the large size of this family and multiple affected members, this pedigree is a unique resource for the detection of the susceptibility gene in Paget's disease. The first susceptibility loci for Paget's disease of bone have been mapped by other investigators to chromosome 6p21 (PDB1) and 18q21.1-q22 (PDB2) in different pedigrees. Linkage analysis of the Australian pedigree in these studies was performed with markers at PDB1: these data showed significant exclusion of linkage, with LOD scores < - 2 in this region (Chapter 4). Linkage analysis of microsatellite markers from the PDB2 region excluded linkage with this region also, with a 30 cM exclusion region (LOD score < -2.0) centred on D18S42 (Chapter 4). This locus on chromosome 18q21.1-q22 contains a serine protease (serpin) cluster with similarities to chromosome 6p21. Linkage analysis of this region also failed to provide evidence of linkage to this locus (Chapter 4). These data are consistent with genetic heterogeneity of Paget's disease of bone. A gene essential for osteoclast formation encoding receptor activator of nuclear factor-kB (RANK), TNFRSF11A, has been previously mapped to the PDB2 region. Mutations in the TNFRSF11A gene have been identified segregating in pedigrees with Familial Expansile Osteolysis and early onset familial Paget's disease, however, linkage studies and mutation screening have excluded the involvement of RANK in the majority of Paget's disease patients. For the Australian pedigree, mutation screening at the TNFRSF11A locus revealed no mutations segregating with affected individuals with Paget's disease (Chapter 4). Based on these findings, our hypothesis is that a novel susceptibility gene relevant to the pathogenesis of Paget's disease of bone lies elsewhere in the genome in the affected members of this pedigree; this gene should be identifiable using a microsatellite genome-wide scan followed by positional cloning. A genome-wide scan of the Australian pedigree was carried out, followed by fine mapping and multipoint analysis in regions of interest (Chapter 5). The peak 2-point LOD scores from the genome-wide scan were LOD = 2.75 at D7S507 and LOD = 1.76 at D18S70. Two additional regions were also considered for fine mapping: chromosome 19p11-q13.1 with a LOD of 1.58 and chromosome 5q35-qter with a LOD of 1.57. Multipoint and haplotype analysis of markers flanking D7S507 did not support linkage to this region (Chapter 5). Similarly, fine mapping of chromosome 19p11-q13.1 failed to support linkage to this region (Chapter 5). Linkage analysis with additional markers in the region on chromosome 5q35-qter revealed a peak multipoint LOD score of 6.77 (Chapter 5). A distinct haplotype was shown to segregate with all members of the family, except the offspring of III-5 and III-6. Haplotype analysis of markers flanking D18S70 demonstrated a haplotype segregating with Paget's disease in a large sub-pedigree (descendants of III-3 and III-4) (Chapter 5). This sub-pedigree had a significantly lower age at diagnosis than the rest of the pedigree (51.2 + 8.5 vs. 64.2 + 9.7 years, p = 0.0012). Linkage analysis of this sub-pedigree demonstrated a peak two-point LOD score of 4.23 at marker D18S1390 (q = 0.00), and a peak multipoint LOD score of 4.71, at marker D18S70. An implication of these data is that 18q23 harbours a novel modifier gene for reducing the age of onset of Paget's disease of bone. A number of candidate Paget's genes have previously been identified on chromosome 18q23, including the nuclear factor of activated T cells (NFATc1), membrane-associated guanylated kinase (MAGUK) and a zinc finger protein. Candidate gene sequencing of these genes in these studies has failed to identify mutations segregating with affected family members in the sub-pedigree linked to chromosome 18q23 (Chapter 6). More recently, a mutation in the gene encoding the ubiquitin-binding protein sequestosome 1 (SQSTM/p62) has been shown to segregate with affected members of Paget's disease families of French-Canadian origin. In this study, a single base pair deletion (1215delC) was identified as segregating with the majority of affected members in the pedigree (Chapter 6). This deletion introduces a stop codon at amino acid position 392 which potentially results in early termination of the protein and loss of the ubiquitin binding domain. The three affected members of the family that do not share the affected haplotype do not carry a mutation in the coding region of SQSTM/p62. Screening of affected members from 10 further Paget's disease families identified the previously reported P392L mutation in 2 (20%) families. No SQSTM1/p62 coding mutations have been found in the remaining 8 families or in 113 aged matched controls. In conclusion, this project has identified genetic loci and mutations that segregate with individuals affected with Paget's disease. Further investigation of the functional significance of the genetic changes at these loci is expected to lead to a better understanding of the molecular basis of this disease.

Migraine, comprised of migraine with aura (MA) and migraine without aura (MO), is a painful neurovascular disease, affecting approximately 16% of the general population. It is characterised by a wide variety of symptoms including headache, nausea and vomiting, and photo- and phonophobia. The disorder is complex involving not only multiple genes, but also specific environmental factors, which can induce attacks in genetically predisposed individuals. Hyperhomocysteinaemia is a known risk factor for cerebrovascular, peripheral vascular and coronary heart disease. The Methylenetetrahydrofolate Reductase (MTHFR) enzyme is involved in homocysteine metabolism. Furthermore, it has been reported that a homozygous mutation (677C to T; Ala to Val) in the 5,10-MTHFR gene is associated with an elevation in plasma homocysteine levels (Frosst et al., 1995). This common mutation in the MTHFR gene has recently been associated with migraine with aura in a Japanese cohort (Kowa et al., 2000). The present study was designed to determine the prevalence of the MTHFR C677T mutation in Australian patients with migraine and to determine whether this mutation is associated with the disease in Caucasians. A large case-control study, consisting of 270 patients with migraine (167 with aura and 103 without aura), and 270 normal matched controls was investigated. Genotypic results indicated that the prevalence of the homozygous (T/T) genotype in migraine sufferers (15%) was higher than that of controls (9%) (P = 0.084). Furthermore, the frequency of the mutant (T/T) genotype in individuals with MA (19%) was significantly higher than in controls (9%) (P = 0.006). Interestingly, the risk of MA was ~2.5-fold higher in suffers possessing the homozygous variant (OR = 2.52, CI: 1.42 - 4.47, P = 0.0012). To confirm the MTHFR allelic association with MA, family-based tests were performed in an independent pedigrees group, where only those with MA were considered affected. Results from both the Pedigree Disequilibrium Test (PDT) and Family-Based Association Test (FBAT) analysis revealed slight, although not significant (PDT test, P = 132; and FBAT test, P = 0.390), over-transmission of the mutant allele (T) from parents to affected offspring. However, despite the MTHFR variant having a high heterozygosity (0.48), there were a limited number of informative transmissions for the MTHFR variant in the pedigree group resulting in reduced power for these tests. In conclusion, our results support the trends reported in the Japanese migraine study and suggest that the homozygous 677T gene variant causing mild hyperhomocysteinaemia, is a genetic risk factor for migraine, and indicate that further studies investigating the role of this gene are warranted. Mutations in various ion channel genes are responsible for neurovascular and other neurological disorders. Inherited ion channel mutations or "channelopathies" are increasingly found to be the cause of various neurological disorders in humans. Wittekindt and colleagues (1998) reported that the calcium-activated potassium channel (hKCa3) gene is a good candidate for schizophrenia and bipolar disorder (BD), as well as for other neurological disorders such as migraine. The hKCa3 gene is a neuronal small conductance calcium-activated potassium channel, which contains a polyglutamine tract, encoded by a polymorphic CAG repeat in the gene. The hKCa3 gene encodes a protein of 731 amino acids containing two adjacent polyglutamine sequences in its N-terminal domain separated by 25 amino acids. The C-terminal polyglutamine sequence is highly polymorphic in length (Austin et al., 1999). hKCa3 plays a critical role in determining the firing pattern of neurons via the generation of slow after-polarization pulses and the regulation of intracellular calcium channels (Kohler et al., 1996). Three distinct mutations in the a1 calcium channel gene have been shown to cause SCA-6, episodic ataxia-2 and familial hemiplegic migraine (FHM) (Ophoff et al., 1996). The hKCa3 gene contains a highly polymorphic CAG repeat that was initially mapped (Chandy et al., 1997) to a schizophrenia locus on chromosome 22 (Pulver et al., 1994). Recently Austin et al (1999) re-mapped hKCa3 and found it to reside on chromosome 1q21, a region that has been linked to FHM (Austin et al., 1999), a rare subtype of MA (Ducros et al., 1997; Gardner et al., 1998), and a region recently showing genetic linkage to typical migraine (Lea et al., 2002). The hKCa3 polymorphism results in small variations in polyglutamine number, similar to those that occur in the calcium channel a1a subunit gene (CACNA1A), which is encoded by CAG expansions and thought to cause Spinocerebellar Ataxia Type 6 via loss of channel function (Austin et al., 1999). Given the recent linkage of FHM to the region of chromosome 1q21, to which hKCa3 resides, and also linkage of typical migraine to this region, a large case-control study investigating this hKCa3 CAG marker and consisting of 270 migraine and 270 stringently matched healthy controls was undertaken. Our results indicated that there was no statistically significant difference in allele distributions for this marker between migraine and non-migraine patients (P >0.05). No significant difference in the allelic distribution was observed in the MA or MO groups when compared to controls (P >0.05) and there was no significant difference in CAG repeat length distribution between the migraine group and controls (P = 0.92), or between the MA and MO groups (P = 0.72) collectively. Hence, the CAG repeat in this gene does not show expansion in migraine. Overall, our results provide no genetic evidence to suggest that the hKCa3 CAG repeat polymorphism is involved in migraine aetiology in Australian Caucasians. Thus the involvement of the hKCa3 gene in migraine is not likely, although the hKCa3 gene remains an important candidate for other neurological disorders that may be linked to the 1q21.3 chromosomal region.

The initial focus of this PhD project was on comparative gene mapping. Comparative gene mapping is facilitated by consensus PCR primers which amplify homologous gene fragments in many species. As a part of an international co-ordinated programme of comparative mapping in pigs, 47 CATS (Comparative Anchor Tagged Sequence) consensus primer pairs for loci located on human chromosomes 9, 10, 20, and 22, were used for amplifying homologous loci in pigs. After optimization of PCR conditions, 23 CATS products have confirmed by comparison with homologous sequences in GenBank. A French somatic cell hybrid panel was used to physically map the 6 porcine CATS products distinguishable from rodent background product, namely ADRA1A, ADRA2A, ARSA, GNAS1, OXT and TOP1. Of these, the map location of ADRA1A and OXT showed inconsistency with the previously recognised conserved relationship between human and pig. The other four loci mapped to positions consistent with known syntenic relationships. Despite low levels of polymorphism, frequently indistinguishable rodent and porcine products in somatic hybrids and some confusion of identity of gene family members, these CATS primers have made a useful contribution to the porcine-human comparative map. The focus of the project then changed to genetic and molecular characterisation of endogenous retroviruses in pigs and their relatives. Pigs are regarded as a potentially good source of organs and tissues for transplantation into humans. However, porcine endogenous retroviruses have emerged as a possible problem as they can infect cultured human cells. Two main types of pig retrovirus, determined by envelope protein, PERV-A and PERV-B, are widely distributed in different pig breeds and a third less common type, PERV-C, has also been recognised. Endogenous retroviruses were analyzed from the Westran (Westmead transplantation) inbred line of pig, specially bred for biomedical research. Thirty-one 1.8 kb env PCR product clones were sequenced after preliminary screening with the restriction enzymes KpnI and MboI. Five recombinant clones between A and B were identified. 55% of clones (17/31) sequenced had stop codons within the envelope protein-encoding region, which would prevent the retrovirus from making full-length envelope protein recognizable by cell-surface receptors of the virus. The endogenous viruses were physically mapped in Westran pigs by FISH (Fluorescence In Situ Hybridisation) using PERV-A and PERV-B envelope clones as probes. Preliminary FISH data suggest that there are at least 22 PERVs (13 PERV-A and 9 PERV-B) and the chromosomal locations of these in the Westran strain are quite different from European Large White pigs. The sequences and mapping results of inbred Westran pig suggest that there are relatively few PERV integration sites compared with commercial pigs and further that a large proportion of clones are defective due to premature stop codons in the envelope gene. To investigate the relationship of endogenous retroviruses in peccaries and pigs, a set of degenerate primers was used to amplify peccary retroviral sequences. The sequences of two putative retroviral clones showed close homology, albeit with a 534 bp deletion, to mouse and pig retroviral sequences. Also, four non-target sequences were amplified from peccary with the degenerate retroviral primers. They are a part of the peccary cofilin gene, a SINE, and a sequence containing a microsatellite. The peccary endogenous retroviral sequences are significant in that they are the first such sequences reported in peccary species and repudiate old claims in the literature that peccaries do not have C-type retroviral sequences.

"July, 2003" / "Amendments of the thesis" and "abbreviations (additional)" inside back cover. / Includes bibliographical references (leaves 195-231) / x, 231, [20] leaves : ill., plates ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Thesis (Ph.D.)--University of Adelaide, Dept. of Paediatrics, 2003

Originally presented as the author's Thesis (doctoral--Universität Hamburg, 1994). / Includes bibliographical references (p. 280-298).

Genetic variation exists in <i>Raphanus</i> that could be of use to <i>Brassica</i> breeders. Of particular interest is the Ogura system of cytoplasmic male sterility (CMS) which has been worked on extensively in a <i>Brassica napus</i> background. Problems have been experienced in <i>B. napus</i>restorer lines due to the inheritance of a large segment of <i>Raphanus</i> chromosome containing the fertility restoring locus. This restorer introgression is located on the <i>Brassica</i> C genome making it only of use for <i>B. napus</i> and not for <i>B. rapa</i> or <i>B. juncea</i>. This thesis describes the development of the materials necessary for the introgression into the <i>Brassica</i> A genome of a defined segment of <i>Raphanus</i> chromosome containing a restorer locus. Defined genetic stocks of <i>Raphanus</i> were developed that contained specific loci controlling restoration of Ogura CMS. This material was used to develop populations segregating for specific restorer loci. Extensive RFLP maps of three <i>Raphanus</i> populations were developed and aligned, resulting in a robust consensus map of the entire <i>Raphanus</i> genome. Three restorer loci were accurately mapped on three separate linkage groups. The segment of <i>Raphanus</i> that is implicated in the restoration of Ogura CMS in a <i>B. napus</i> restorer line developed by INRA was identified and it did not correspond to any of the regions containing the three mapped restorer loci, suggesting the presence of more restorer loci in <i>Raphanus</i>. Comparative mapping between the <i>Raphanus</i> genome map and previously generated <i>Brassica</i> A genome RFLP maps demonstrated large regions of collinearity between segments of chromosomes of the two species. Preliminary examination of the two genome maps suggest they contain essentially the same overall genetic content but with large segments of the genomes rearranged with respect to each other. Likely sites of <i>Raphanus</i> restorer introgression into the <i>Brassica</i> A genome were predicted. Trigenomic tetraploids were developed in which pairing and recombination between homoeologous segments of <i>Raphanus</i> and <i>Brassica</i> A chromosomes should result. Progeny of these individuals will allow an assessment of the pattern and extent of recombination that occurs between the chromosomes of the <i>Raphanus</i> and <i>Brassica</i> A genomes and should lead to the development of 'B. napus' lines carrying Ogura CMS restorer alleles from <i>Raphanus</i>.

Thesis (PhD)--Macquarie University, Division of Environmental and Life Sciences, Department of Biological Sciences, 2002. / "November 2001". Bibliography: leaves 136-157.

Rationale. Definition of complex behavioral disorders is generally phenomenological in nature and guided by pragmatic, rather than genetic, concerns. Consequently, important aspect of genetic analysis is the search for novel phenotypic definitions from the familial/genetic perspective. SLI study 1. SLI denotes an inability to acquire normal language in the absence of peripheral hearing impairment, neurological disorder, and mental retardation. Sibling resemblance for several theoretically derived specific components of the SLI phenotype was examined in families of SLI children. In 38 sib-pairs from 10 French-speaking pedigrees, Verb Tense Morphology sub-tests (Real and Non-real Words) showed nonparametric correlations of 0.39 and 0.35, respectively (p < 0.05, 2-tailed). In a densely affected Anglophone pedigree, 41 sib-pair showed familial resemblance with respect to Derivational Morphology (r = 0.52, p < 0.01). SLI study 2. Family history study in 27 families examined the relationship between attention deficit/hyperactivity in SLI children and familial risk of speech/language disorders. Higher odds of speech/language disorders were observed in first-degree relatives of 13 SLI children who also had a medical record of attention deficit/hyperactivity (15/27 vs. 4/46, p = 0.001). Alcoholism study 1. Latent class analysis (LCA) including gender and 15 antisocial behaviors (>15yr) was performed in 236 broadly ascertained alcohol-dependent subjects (121 males, 115 females). Evidence for 3 qualitative behavioral classes was obtained: Socially Adjusted Adults, SAA; Antisocial Non-Aggressive Adults, ANAA; and Antisocial Aggressive Adults, AAA. In both, genders, the AAA class had the earliest age of onset for alcohol dependence (p = 0.001), more alcoholic first-degree relatives and more of other psychopathology. In females, the ANAA class was intermediate. In the ANAA males, socially adjusted childhood behavior differentiated the late onset from the intermediate ons

The objective of the present study was to investigate the potential use of bulk segregant analysis (Michelmore et al., 1991) as a method to rapidly identify genetic markers linked to traits in sugarcane. Four bulked DNA samples were prepared from progeny of a sugarcane cross, AA157, based on segregation for the fibre trait. The bulks comprised five and ten individuals on either side of the fibre phenotypic extreme. The random amplified polymorphic DNA (RAPD) technique (Williams et ai., 1990) was used to screen for differences between the low and high fibre bulks. A total of 749 fragments were amplified in the bulks, eight of which were polymorphic. The segregation of the bulk specific polymorphism was analysed in 80 progeny of the AA157 cross; and seven were found to reproducibly segregate on a 1: 1 basis. This indicates that they are single dose fragments. A total of 79 polymorphisms were detected between the parents of the cross, indicating 10.5% variation in the genomic region sampled. Twenty two of the parental polymorphisms segregated as single dose fragments in the progeny of the cross AA157. Analyses of variance (ANOVAs), and multiple regression analyses, were used to ascertain linkage of the putative RAPD markers to fibre, and if linked, to determine the fibre variation ascribed respectively. Three RAPD fragments were found linked to the fibre trait. Fragments OPA17438 and OPC16889 (at the 99% significance level), and OPB1l464 (at the 95% significance level). These putative markers ascribed a total of 28.6% fibre variation in the 1993 season. The association of the RAPD markers with fibre in the different seasons (1992, 1993 and 1994) was investigated. Three RAPD markers were found linked to the fibre trait in each season, with a total of 5.5% and 31,4% fibre variation ascribed in the 1992 and 1994 seasons respectively. Marker OPA17438 was found to be linked to the fibre trait in all three seasons investigated, and marker OPC16889, was found linked to the fibre trait in the 1992 and 1993 seasons. Cross validation of the linkages of the RAPD markers to the fibre trait was carried out by a modified form of 'jacknifing' where the sample size was reduced to N-l0, and RAPD marker-fibre trait associations investigated as before. RAPD markers OPA17438 and OPC16889 were still consistent across the seasons, however marker OPA17438 was no longer linked to the fibre trait in the 1992 season. To investigate the genetic behaviour of RAPD based markers in sugarcane and the potential for their application in marker-assisted selection (MAS), two putative RAPD markers were converted to sequence characterised amplified regions (SCARs) (Paran and Michelmore, 1993). The RAPD fragments OPA17438, OPBl1464, and OPC16889 were excised from agarose gels, re-amplified and cloned into the pCR-Script SK (+) phagemid for sequencing. RAPD markers OPA17438 and OPB11 464 were converted to SCARs by using their sequences to design longer specific primers. A third SCAR marker, SAl1640, originally derived from sugarcane cDNA as a potential stem preferential expressed sequence tag, was included in the analysis to increase the sample size. All three SCAR markers segregated in a monomorphic fashion in the parents and progeny of the cross AA157. In addition, monomorphic length variants for markers, OPA17438 and OPB11 464 were detected with the SCAR amplification. All three SCARs segregated in a monomorphic fashion in different commercial varieties and bulks of S. officinarum and S. spontaneum, the progenitors of modern commercial varieties. The segregation analyses of the SCAR markers indicate that the RAPD polymorphism of marker SAl1640 was probably due to a point mutation or mismatch in the priming site. The segregation analyses of SCARs for the markers OPA17438 and OPB11464 indicate that their segregation in the RAPD analyses was due to an insertion mutation in the genetic locus. The combined results of the SCAR and RAPD segregation of markers OPA17438 and OPB11464 are indicative of preferential pairing in the cross AA157. Finally, to investigate the extent of linkage disequilibrium in a modern commercial variety, twenty two single dose RAPD fragments were investigated for their association with four traits in 53 progeny of cross AA157. The four traits investigated were fibre %cane, brix %cane, pol %cane and ers %cane over three seasons (1992, 1993 and 1994), at different ages of harvest (12, 8, and 9 months respectively). Seventeen linkages of RAPD markers to the four traits, over the three seasons, were detected. The phenotypic variation ascribed by the RAPD markers ranged from 7.6% fibre %cane variation explained by one marker in 1992, 29.6% fibre %cane (three markers) in the 1993 season to 10% (three markers) in 1994. A total of 14.1% brix %cane variation was ascribed by two markers in 1992, 9.6% (one marker) in 1993 and 16.3% (two markers) in the 1994 season. A total of 13.5% estimated recoverable sucrose %cane was ascribed by one marker in 1992, 12% (two markers) in 1993 and 15.3% (two markers) in the 1994 season. Two markers explained 17.2% pol %cane variation in 1992 and 25.4% in the 1994 season. Only four markers were detected across different environments, three of which were linked to fibre. These were OPA17438, OPB16618 and OPC16889, each linked to fibre in two seasons. RAPD marker OPB11 464 was linked to estimated recoverable sucrose %cane in two seasons. Two markers were found associated with different traits in a single season. RAPD marker OPB11 464 was found associated with brix %cane and estimated recoverable sucrose %cane in the 1993 season, and RAPD marker OPA17438 was found associated with all four traits in the 1994 season. / Thesis (Ph.D.)-University of Natal, Durban, 1998.