It is well established that polymorphisms of immune-related genes modulate host susceptibility to autoimmune diseases, including MS and SLE [27
]. Historically, most studies have focused on polymorphisms that result in the replacement of amino acids [27
]. More recently, substantial information has been accumulated that demonstrates that polymorphisms at the promoter and intron regions can also have a significant impact on susceptibility. These alterations modulate either RNA synthesis (transcription) or splicing [41
]. Although it is well established that the 3′ UTR plays a major role in RNA stability, we are not aware of any study reporting that polymorphism at the 3′ UTR modulates susceptibility to autoimmune diseases by changing mRNA stability. Our data presented in this study revealed that a destabilizing dinucleotide deletion in the 3′ UTR of the CD24 gene may confer a significant protection against the risk and progression of MS and against the risk of SLE. Our conclusion is based on five lines of evidence.
First, a population study with 275 independent Caucasian MS patients and a comparable size of normal controls revealed that individuals with the deletion in at least one allele had about a 2-fold less relative risk in comparison to those without the deletion. Thus, the CD24 P1527del
allele may be a protective genetic susceptibility factor for the onset of MS. This is more remarkable in light of the fact that polymorphisms at sites that were only 100–500 bp apart did not have a significant impact on the risk of MS. The strong association at P1527, but not at the nearby SNPs, suggests that the deletion was causatively related to the reduced MS susceptibility. This interpretation is consistent with the fact that the frequencies of the associated alleles at the two nearby (flanking) loci are very different from that of the protective allele. A recent study showed that the power to detect the association in such loci is diminished even when there is high linkage disequilibrium [43
]. This also leads to a reasonable explanation as to why two loci in high LD are not both associated with the disease.
Second, using data from two independent cohorts of families, we also established a strong association of the CD24 P1527 polymorphism with MS. The P1527TG
allele was preferentially transmitted to affected individuals. This result strongly supports the conclusion from the case-control analysis that the P1527del
allele may be a protective genetic susceptibility factor for the onset of MS. Both of these results remain significant after multiple-testing adjustments. Within the Ohio State University (OSU) cohort, our previous data revealed that the P170T
allele was preferentially transmitted into affected individuals among multiplex families with two or more MS patients [27
]. This result continues to hold with our expanded OSU family set, although not with the MSGG cohort (data not shown). In summary, results from both of population and the family studies confirm our earlier conclusion that the CD24
locus is a major modulator for MS risk.
Third, survival analysis revealed a significant association (even after correcting for multiple tests) of CD24 P1527 with MS disease progression; MS patients with the P1527del allele had a significantly delayed progression. This finding further confirms that the P1527del allele is a protective genetic factor for MS. An interesting issue is whether P1527 is associated with the progression of MS because of its linkage to P170. We consider it very unlikely as our analysis of LD revealed that there is little LD between the two sites despite their close proximity to one another, perhaps due to a recombination hotspot within the CD24 gene. Moreover, P1056, which is closer to P170, is not associated with the progression of MS. We therefore consider it likely that P170 and P1527 are independently associated with the progression of MS. Since P1626 is less than 100 bp away and shows a strong LD with P1527, it remains possible that its association with MS progression may be due to its proximity to P1527. This interpretation is favored as P1626 shows no association with MS risk. Since our analysis has now covered all known CD24 polymorphisms in the exons, it is likely that P1527, rather than other SNPs, is related to protection against autoimmune diseases.
Fourth, in addition to MS, which is an organ-specific autoimmune disease, we also observed that the CD24
allele is preferentially transmitted to unaffected individuals in the SLE family data. It is worth noting that the SLE data should not be regarded as a replication of MS data per se. Rather, our data suggest that the protective effect of the dinucleotide deletion extends to systemic autoimmune diseases. Thus, in addition to its critical role for T-cell proliferation in the central nervous system [25
], CD24 may play a role in the development of multiple autoimmune diseases.
Based on the observed data pattern and the structure of the family cohorts, we have chosen TRANSMIT soft ware to detect association between CD24 polymorphism and risk of autoimmune diseases to maximize the statistical power. However, we caution that TRANSMIT may have inflated type-I error due to its inferences of missing parental genotypes [44
]. Nevertheless, we do not believe the core finding is due to type-I errors, as statistically significant association can also be find with FBAT that deletes data from families without parental information (MS dataset, p
= 0.04; SLE data set, p
Fifth, the dinucleotide deletion reduced steady levels of CD24 mRNA by more than 2-fold. Thus, in heterozygous patients, the mRNA from the alleles with the deletion was only 50% of that of the alleles without the deletion. This is recapitulated in transfection studies. Analysis of RNA decay kinetics revealed that the half life for the CD24 transcript with the dinucleotide deletion was at least 4-fold shorter than that of the wild-type allele. Since CD24 was expressed at high levels among some lineages of hematopoietic cells and in the transfected CHO cells, the reduction in the steady levels may underestimate the reduction in other cell types, such as T cells, in which CD24 is expressed at lower levels and is therefore less likely to saturate the degradation system. The low expression of CD24 in T cells is essential for homeostatic proliferation of T cells, which has been implicated in the development of autoimmune diseases.
In summary, we demonstrated that a dinucleotide deletion at the 3′ UTR of the human CD24
gene confers significant protection against the risk and progression of MS and the risk of SLE. These results not only provide insight into the genetic basis of MS and SLE susceptibility, but, perhaps more importantly, to our knowledge, this is the first report that shows how polymorphisms at the 3′ UTR modulate susceptibility to autoimmune diseases by regulating RNA stability. Since CD24 is a checkpoint for homeostatic proliferation of T cells [29
], which is implicated in other autoimmune diseases [45
], it will be of great interest to test the contribution of CD24 to the risk and progression of other autoimmune diseases.