Evidence of Interaction between PTPN22 and p53 codon 72 Polymorphisms on Susceptibility to Immune Related Diseases

This work was carried out in collaboration between all authors. Authors FGB, EB, AM and GR designed the study, wrote the protocol and the first draft of the manuscript. Authors PS, NR, AN and LC managed the control of laboratory methods, collection of data, genotype determination, preparation of data files and statistical analysis. Authors FGB, AM, PS and AN managed the literature search. All authors read and approved the final manuscript. ABSTRACT Background: PTPN22 codifies for a protein-tyrosine-phosphatase (Lyp) involved in T cell receptor signaling regulation. p53 is involved in immune related inflammation regulating STAT 1 and pro-inflammatory cytokines. Possible interaction between the two systems concerning the susceptibility to immune related disorders are therefore biologically plausible. In the present note we have searched for such interaction in type 1 carriers does not differ significantly from controls while in subjects carrying the *Pro allele is higher in patients than in controls. A significant increase of Odds Ratio is observed only in presence of both *T and *Pro alleles suggesting a cooperative interaction. Conclusion: It has been suggested that the susceptibility to autoimmune disorders in the presence of *T allele could be related to failure to delete auto reactive T cell during intrathymic selection. *Pro allele variant with its strong transcriptional activity could enhance the multiplication of such auto reactive T cell escaping intrathymic thus explaining a significant increase of Odds Ratio in the presence of both factors .The present observation could have relevance to identify individuals at high risk of clinical manifestations.

carriers does not differ significantly from controls while in subjects carrying the *Pro allele is higher in patients than in controls. A significant increase of Odds Ratio is observed only in presence of both *T and *Pro alleles suggesting a cooperative interaction. Conclusion: It has been suggested that the susceptibility to autoimmune disorders in the presence of *T allele could be related to failure to delete auto reactive T cell during intrathymic selection. *Pro allele variant with its strong transcriptional activity could enhance the multiplication of such auto reactive T cell escaping intrathymic thus explaining a significant increase of Odds Ratio in the presence of both factors .The present observation could have relevance to identify individuals at high risk of clinical manifestations.

INTRODUCTION
PTPN22 codifies for a protein-tyrosin-phosphatase (LYP) involved in the regulation of T cell receptor signaling (TCR). The gene shows a single nucleotide polymorphism C/T at +1858 resulting in the W620 variant that is associated with autoimmune disorders [1].
Codon 72 in exon 4 of p53 gene shows a polymorphism due to the substitution of G with C that determines the change of arginine to proline in the protein. The amino acid change affects biochemical and functional properties of p53: the proline variant (*Pro allele) is a stronger transcriptional activator, whereas the arginine variant (*Arg allele) is a stronger apoptosis inducer [2]. Activation of p53 gene has been associated with immune related disorders [3,4] and anti p53 autoantibodies have been found in the serum of patients with these disorders [5].Recent studies suggest that p53 may be involved in immune related inflammation regulating STAT1 and proinflammatory cytokines [6,7]. Possible interactions between the two systems concerning the susceptibility to autoimmune disorders are therefore biologically plausible.
In the population of Rome *T allele of PTPN22 and *Pro allele of p53 codon 72 polymorphisms have been found more frequent in type 1 diabetes (T1D) than in controls [8-10, therefore, we have searched for possible interaction between the two systems on susceptibility to T1D. We have previously observed an interaction between PTPN22 and p53 codon 72 in endometriosis [11] and in Coronary Artery Disease (CAD) [12] in which immunological factors couldy have a significant role: thus, in the present note we have also reviewed our data on these diseases. The pattern of interaction is concordant in these disorders suggesting a significant effect on susceptibility in presence of both *Pro allele of p53 codon 72 and of *T allele of PTPN22.

MATERIALS AND METHODS
We have studied 287 subjects with type 1 diabetes (T1D), 129 non diabetic subjects admitted consecutively to Valmontone Hospital (Italy) for CAD (Coronary Artery Disease), 130 women with endometriosis and 256 healthy blood donors. All subjects were from the White Caucasian population of Rome. Written informed consent was obtained from each subject. The study protocol conforms to the Ethical Guidelines of the 1975 declaration of Helsinki and was approved by the Ethical Committee of Valmontone Hospital. All subjects were collected consecutively and separately for each sample in order to perform a casecontrol study.
The p53 codon 72 polymorphism (chromosome 17) has three genotypes: *Arg/*Arg, *Arg/*Pro and *Pro/*Pro and was evaluated using the restriction fragment length polymorphism polymerase chain reaction method described by de la Calle-Martin et al. [2] as previously described [11.
Chi-square test of independence and odds ratio analyses were performed by the SPSS programs [14]. Three way contingency table analysis by a log linear model was carried out according to Sokal and Rohlf [15.

RESULTS
Clinical data of the patients and controls are shown in Table 1. Table 2 shows the proportion of *T allele carriers in *Arg/*Arg genotype and in *Pro allele carriers of p53 codon 72 polymorphism in healthy controls and in subjects with immune related diseases. In all these disorders but not in controls the proportion of *T allele is higher in carriers of *Pro allele than in *Arg/*Arg genotype. The proportion of *T allele carriers in *Arg/*Arg subjects with immune related disorders does not differ significantly from controls but in carriers of *Pro allele is significantly higher in patients than in controls (p=0.0005). Table 3 shows the proportion of subjects carrying both *T allele of PTPN22 and *Pro allele of p53 codon 72 in diseases and in controls. No significant difference is observed among the three diseases. The proportion of subjects carrying both *T and *Pro alleles is significantly higher in the diseases than in controls (Odds Ratio=3.736; 95%C.I. 1.506-9.853; p=0.002). Table 2 has shown a lack of a significant three way interaction and a significant additive effect of PTPN22 and p53 codon 72 on susceptibility to diseases considered (p=0.003,data not shown). These results appear against the hypothesis of epistasis and in favor of cooperative interaction.

A three way contingency table analysis by a log linear model performed on 2x2 contingency table reported in sections (A) and (D) of
Odds ratio analysis is also in favour of an additive effect of the two alleles on susceptibility to the diseases. Fig. 1 shows the Odds Ratio in relation to presence of one risk factor only (*Pro allele or *T allele) and to presence of both risk factors (*Pro and *T alleles). A very high Odds Ratio is observed in presence of both factors.
In T1D the pattern of associations described in Tables 2 and 3 does not depend neither on sex nor on age at onset. In CAD the pattern is not influenced by age. In endometriosis diffusion of lesion, parity and duration of pharmacological treatment do not influence significantly the pattern of associations shown in Tables 2 and 3. Significant differences have been observed concerning the socio-economic status with a higher proportion of the high risk joint genotype in the low as compared to the high status (p<0.05).

DISCUSSION
Our observations suggest that the susceptibility to the diseases studied is increased in presence of both *Pro and *T alleles. It has been suggested that the susceptibility to autoimmune disorders in presence of *T allele of PTPN22 could be related to failure to delete auto reactive T cell during intratymic selection [16]. *Pro allele variant of p53 codon 72 with its strong transcriptional activity could enhance the multiplication of auto reactive T cell escaping intratymic selection contributing to emergence of clinical manifestation. This could explain the high risk in presence of both factors.
The limitation of the present study is represented by the relatively small number of subjects examined. Therefore our observations need to be confirmed in an independent clinical setting. However the concordance of the pattern observed in the three diseases makes unlikely the possibility of a mere chance sampling artifact. It would be interesting to search for such interaction also in other diseases involving the immune system.
Appropriate studies on immunological parameter in presence of only one and in presence of both risk factors could contribute to elucidate the mechanism of interaction. From a practical point of view the observation could have relevance to identify individuals at high risk of clinical manifestations. It would be also interesting to elucidate the possible role of these polymorphic genes of the clinical evolution of the diseases: unfortunately we have no reliable information of this important aspect of the problem.

CONCLUSION
The present data suggest a cooperative interaction between PTPN22 and P53 codon 72 polymorphisms conserning their effects on susceptibility to immune related diseases.
Genetic analysis of multifactorial disorders represents an important problem for medical genetics: the study of single genetic factors in a mendelian perspective is reductionist and cannot solve the problem. It is likely that simultaneous analysis of multiple genes related to a specific function will provide a more productive approach to clarify the etiology of multifactorial diseases.