Prevalence of the serpin peptidase inhibitor ( alpha-1-antitrypsin ) PI * S and PI * Z alleles in Brazilian children with liver disease

Serpin peptidase inhibitor, clade A (alpha-1 antiproteinase, antitrypsin), member 1 (SERPINA1) deficiency is one of the main genetic causes related to liver disease in children. In SERPINA1 deficiency the most frequent SERPINA1 alleles found are the PI*S and PI*Z alleles. We used the polymerase chain reaction and the amplification created restriction site (ACRS) technique to investigate the prevalence of the PI*S and PI*Z alleles in a group of Brazilian children (n = 200) with liver disease and established the general frequency of the PI*S allele in our population. We found a significant association of the PI*Z allele and liver disease, but no such relationship was found for the PI*S allele. Our results show that SERPINA1 deficiency due to the PI*Z allele, even when heterozygous, is a frequent cause of liver disease in our group of Brazilian children but that the PI*S allele does not confer an increased risk of hepatic disorders in our group of Brazilian children

Serpin peptidase inhibitor, clade A (alpha-1 antiproteinase, antitrypsin), member 1 (SERPINA1, formerly known as alpha-1-antitrypsin (A1AT) is a glycoprotein mainly produced by hepatocytes whose major function is to inhibit the action of neutrophilic elastase, a serine protease that hydrolyses elastin fibers in the lung (Francavilla et al., 2000).The protein is encoded by the highly polymorphic SERPINA1 gene located in the long arm of chromosome 14 (14q31-32.3)(Schroeder et al., 1985).Mutations in SERPINA1 lead to a reduction or loss of the inhibitory capacity.Although more than seventy alleles for the enzyme are known, designated according to their isoelectric point (De Tomasso et al., 2001), the PI*S and PI*Z alleles are the two major forms predominant among the deficient variants.The PI*S allele results from a point mutation in exon III of the gene, which leads to the substitution of a glutamic acid residue for a valine at position 264 and formation of an unstable protein (Long et al., 1984).This is associated with a reduction of about 40% in the normal range of SERPINA1 and increased intracellular degradation (Cox, 2004).Even lower levels of SERPINA1 are produced by the PI*Z allele, this variant resulting from the substitution of a glutamic acid residue for a lysine at position 342 in exon V (Brind et al., 1990) and is the most frequent SERPINA1 deficient variant (Alpha-1 Foundation, 2003).Classical studies demonstrated an association of the PI*Z allele with liver disease (Lieberman et al., 1972).Hepatic involvement in SERPINA1 deficiency due to the PI*Z allele seems to be due to the accumulation of mutant proteins in hepatocytes (Perlmutter, 2003) that have been associated with the clini-cal features of SERPINA1 deficiency in the liver (mainly neonatal cholestasis) that can progress to chronic liver disease and cirrhosis (Sharp et al., 1969;Teckman et al., 1996;Lomas and Parfrey, 2004).Deficiency in SERPINA1 occurs in 1 in 1600 to 1 in 1800 live births, but prospective natural history studies indicate that only 10% to 15% of the affected population develops clinically significant liver disease (Sveger, 1976).Interestingly, the follow up of a cohort of patients with SERPINA1 deficiency showed that only 3% of children who are homozygous for the PI*Z allele develop liver disease (Sveger and Eriksson, 1995).More recently, a possible role of SERPINA1 as a modifier gene of other forms of pediatric liver disease has been suggested (Campbell et al., 2007).
The objective of the study reported in this paper was to evaluate the frequency of the PI*S and PI*Z alleles in pediatric patients with liver disease from different regions of Brazil and to compare these data with a control group of anonymous blood donors to establish the general frequency of the PI*S allele in our population.
We investigated DNA samples from 200 children (52% male, aged one month to 12 years with a median age of one year) with liver disease for SERPINA1 mutations E264V (PI*S allele) and E342K (PI*Z allele).The children were referred to the Genetic Therapy Center at the Clinical Hospital in Porto Alegre (Centro de Terapia Gênica, Centro de Pesquisas, Hospital de Clínicas de Porto Alegre (HCPA), Porto Alegre, Rio Grande do Sul, Brazil) by nine hospitals from different regions of Brazil, mostly from the Southern (84.5%), Southeastern (9.5%) and Center-West (1.5%) region.The main reasons for suspicion of SERPINA1 deficiency were cholestatic jaundice (23.5%), cryptogenic cirrhosis (10%) and neonatal cholestasis (10%).Other reasons include elevated transaminases, steatosis, hepatosplenomegaly, hepatic failure and ascites.The diagnosis of cirrhosis was based on clinical, biochemical, ultrasonographic and endoscopic findings and/or histology.Cryptogenic cirrhosis was diagnosed based on the absence of inborn errors of metabolism, sclerosing cholangitis, viral hepatitis B and C, congenital infectious disease, use of hepatotoxic drugs and Wilson disease.Blood samples were taken from the children as part of normal clinical investigations between 1998 and 2006.
The presence of the PI*S allele was also investigated in blood samples from 150 unidentifiable voluntary donors who donated blood at the HCPA blood bank in 1998.These samples had already been tested for the PI*Z allele (Lima et al., 2001) but this time were used as a control to estimate the frequency of the PI*S allele in the general population, the use of adults as the control population being justified by the fact that only a small percentage of individuals with the PI*ZZ allele develop liver disease (Sveger and Eriksson, 1995).
Although we were not able to ethnically classify the children or adults who participated in our study almost 85% of them were from the Brazilian state of Rio Grande do Sul which has an ethnically heterogeneous population with a predominance of individuals of Mediterranean and Central European descent and, compared to the rest of Brazil, a small contribution of genes of African and Amerindian origin (Carvalho-Silva et al., 2001).Informed consent was obtained from the parents of the children and the blood donors prior to their entering the study and the study was approved by the ethics commission of HCPA.
Genomic DNA was extracted from 5 mL of peripheral blood using the salting-out technique (Miller et al., 1988) and DNA analysis performed using the polymerase chain reaction (PCR) and the amplification created restriction site (ACRS) technique (Andresen et al., 1992).To detect the PI*S allele we amplified SERPINA1 exon III in a final volume of 50 μL containing 30 pmol each of primers p7553 (5'-CGTTTAGGCATGAATAACTTCCAGCA-3') and p7702 (5'-GATGATATCGTGGGTGAGTTCATT TA-3'), 5 μL 10X buffer (200 mM Tris-HCl (pH 8.4) and 500 mM KCl), 1.5 mM of MgCl 2 , 0.2 mM of each dNTP and 1 unit of Taq DNA polymerase.Amplification of the PI*Z allele SERPINA1 exon VII used a final volume of 50 μL containing 40 pmol each of primer AT5f (5'-ATAA GGCTGTGCTGACCATCGTC-3') and AT5r (5'-GAAC TTGACCTCGAGGGGGATAGA-3'), 5 μL of buffer (75 mM Tris-HCl (pH 9), 50 mM KCl, 20 mM (NH 4 ) 2 SO 4 , 2 mM MgCl 2 and 0.001% (v/v) bovine serum albumin), 0.2 mM of each dNTP, 6% (v/v) DMSO and 1 unit of Taq DNA polymerase.Amplification was carried out in a Personal Thermocycler (Eppendorf, Germany) using an annealing temperature of 50 °C for both exons.Amplification length was 149 bp for PI*S and 97 bp for PI*Z.The PCR products were cleaved using the using the XmnI endonuclease for the PI*S products and TaqI for the PI*Z products according to manufacturer's instruction (Invitrogen).Fragments were separated on 12% (w/v) polyacrylamide gel and stained with 0.004% (w/v) ethidium bromide.Expected fragment sizes were 133 bp and 16-bp for homozygous PI*S individuals and 111 bp, 22 bp and 16 bp for individuals without PI*S.For PI*Z homozygous individuals the fragments were 86 bp and 11 bp.When the mutation was not present, fragments were 64 bp, 22 bp and 11 bp.All reagents for PCR and digestion were purchased from Invitrogen (Carlsbad, USA).
Statistical analysis to compare the allele frequency in patients and controls was performed using the chi-square test with the Yates correction.
Our analysis showed that of the 200 children examined, 20 (10%) were PI*Z homozygous, 2 (1%) were PI*S homozygous, 2 (1%) were PI*SZ compound heterozygous, 13 (6.5%)were PI*Z heterozygous and 21 (10.5%) were PI*S heterozygous.The HCPA clinical findings for the children positive for the PI*Z and PI*S alleles are available as supplementary online material (Table S1).Cirrhosis was the final diagnosis in 2 children heterozygous for PI*Z and in 1 child heterozygous for PI*S who had associated biliary atresia.Eight PI*ZZ children developed cirrhosis and five were given a liver transplant.The median time of follow-up for these patients was 7 years (range = 2 years to 19 years).One PI*SZ child had Overlap Syndrome and was listed for a liver transplant 6 months after diagnosis with severe decompensate cirrhosis.No liver disease was observed in children homozygous for the PI*S allele.
To obtain data about the general frequency of the PI*S and PI*Z alleles in our population we also investigated 150 blood donors, who acted as controls representing the frequency of these two alleles in the general population.In this group the PI*S allele frequency found was 6.66%.Eighteen (12%) individuals were heterozygous and only one (0.66%) was homozygous for PI*S.The PI*Z allele occurred in only 1 (0.66%) heterozygous individual.
Comparison of the calculated allele frequencies for the children and those for the general population as estimated from the control group revealed a significantly (p < 0.001) increased PI*Z allele (13.75%) frequency for the children with liver disease as compared to controls.Moreover, the association between liver disease and the PI*Z allele was significant (p < 0.01) even considering only heterozygous children.These children were confirmed to not have other changes in SERPINA1 (data not shown).The frequency of the PI*S allele (6.75%) was not statistically different from that found in the general population for homozygous and heterozygous individuals (Table 1).Analysis of genotypic frequencies indicated that the PI*Z allele was in Hardy-Weinberg disequilibrium for the children with liver disease.
Deficiency in SERPINA1 is a genetic disorder strongly related to hepatic disease and the need for liver transplantation.It affects all major racial subgroups, and there are about 120.5 million carriers and deficient individuals worldwide (De Serres et al., 2003).The PI*S allele is a very common deficient variant, reaching incidences higher than 14% in countries such as Portugal and France (Roychoudhury and Nei, 1988).Luisetti and Seersholm (2004) published data on the worldwide PI*S and PI*Z allele frequencies, but presented no data for South America due to the lack of studies in this region.In our study, we analyzed the frequency of the PI*S allele in a sample of Brazilian children with liver disease and also in a group of blood do-nors representative of the general population.Comparing our data with that of previous investigations, the frequency of the PI*S allele in our population seems to be higher than in Holland (2.9%) and Denmark (2.2%) but lower than in Portugal (15%) and France (14.5%) (Roychoudhury and Nei, 1988;Luisetti and Seersholm, 2004).The intermediate frequency of the allele in our Brazilian population may be explained by the mixed origin of our population, which has a strong influence from Portuguese colonization but with a significant contribution from other Central European countries, as observed by mutation frequency data for other diseases (Castro et al., 2007).
In our study, the PI*Z allele showed a high prevalence among the children, strengthening the hypothesis that the presence of the PI*Z allele contributes to hepatic disease and may have a role as a modifier gene of other forms of pediatric liver disease (Campbell et al., 2007).However, it is important to point out the lack of correlation between hepatic disease and the presence of the PI*S allele seen in our investigation.Elliot et al. (1996) suggest that the PI*S variant has increased susceptibility to polymerization, although this increase is marginal when compared to the PI*Z allele.There has been some speculation regarding synergy between the variants which may lead to cirrhosis in PI*SZ adults (Mahadeva et al., 1999).Hadzic et al. (2005) investigated PI*SZ patients and showed that although they had low SERPINA1 levels some of them may have had late hepatic disorders.This suggests that even though we found no correlation between the PI*S allele and hepatic disorders in children, individuals with the PI*S allele should be monitored in adulthood to evaluate late-onset hepatic problems.Furthermore, although testing for both alleles in children with liver disease may not be cost-effective further assessment is needed regarding the cost-effectiveness of testing for the PI*S allele in heterozygous adults with the PI*Z allele.
Summarizing, in this study of SERPINA1 we found a significant association between liver disease in children and the PI*Z allele but no such relationship for the PI*S allele.These results indicate that while SERPINA1 deficiency due to the PI*Z allele is a frequent cause of liver SERPINA1 in children with hepatic disease 425 *Within the same column differences between the values were significant at the stated probabilities (p).ns Within the same column differences between the values were not significant at the stated probabilities (p).
disease in children the PI*S allele does not confer an increased risk for pediatric hepatic disorders.

Table 1 -
Serpin peptidase inhibitor (alpha-1-antitrypsin) SERPINA1 PI*S and PI*Z allele frequencies and homozygous and heterozygous genotype frequencies for Brazilian children (n = 200) with liver disease and adult blood donors (n = 150) with no liver disease compared using the chi-squared test with Yates correction.