Novel Kelnull Allele Detected In a Brazilian Woman with the Kell Null Phenotype.

Abstract 1111

Kell is the most important blood group system after ABO and Rh because all frequently occurring Kell-specific antibodies must be considered clinically significant. The KEL antigens are carried by the human red cell (RBC) membrane Kell glycoprotein (CD238), a proteolytic enzyme encoded by the 19-exon KEL gene on the long arm of chromosome 7 (7q33). Kell null (K₀) is a very rare phenotype characterized by the absence of the Kell protein and all KEL antigens on RBC surface caused by different molecular defects on KEL gene. K₀ persons, however, may produce anti-KEL5 (anti-Ku) antibody that is associated to severe hemolytic transfusion reaction and perinatal anemia. Although the K₀ phenotype has been known since 1957, the first reports of the K₀ molecular basis occurred only in 2001. At present, 23 KELnull alleles are recognized to abolish the KEL antigens expression, most of them due to stop codon and alternative splicing caused by single nucleotide mutation. In this study, we report a novel KELnull allele detected in a Brazilian woman with the K₀ phenotype. We investigated a 59-year-old Caucasian-Amerindian descent woman (proband) from Northeast Brazil who showed an antibody that reacted against high-prevalence RBC antigens during a pre-transfusion evaluation for orthopedic surgery. We also were able to evaluated three relatives (proband′s sister and two nieces) and normal RBC controls. We performed standard immunohematological methods and PCR-RFLP KEL genotyping as screening tests. To identify the molecular defect, we used sequencing technique covering all 19 coding and exons-introns regions of the KEL gene. To confirm the K₀ phenotype and exclude KELel (K_mod) phenotype we employed flow cytometry, a higher sensitive technique, using soroclone anti-KEL1(anti-K), anti-KEL2(anti-k), anti-KEL3(anti-Kp^a), anti-KEL4(anti-Kp^b) and anti-CD238 antibodies. Proband′s frozen RBCs were washed 3x in 3ml PBS and 2×10⁶/50μl/tube were incubated for 30min at room temperature with monoclonal antibodies (anti-K, -k, -Kp^a, -Kp^b – 50μl; anti-CD238 1:200,10μl). FITC conjugated polyclonal rabbit anti-human Ig and goat-anti-mouse Ig were used as second antibodies. Data acquisition and analysis in 20,000 events at RBC region were performed using FACscalibur flow cytometer and CellQuest program. Previously known (A) KEL:-1-,2,-3,4 (K-,k+,Kp^a-,Kp^b+) and (B) KEL:1,2,-3,4 (K+,k+,Kp^a-,Kp^b+) RBCs samples were used as positive controls. Fresh and thawed RBCs stained with PE conjugated anti-human glycophorin were also tested. RBCs of the proband were phenotyped as KEL:-1,-2,-3,-4,-6,-7 (K-,k-,Kp^a-,Kp^b-,Js^a-,Js^b-), and the RBC alloantibody was identified as anti-KEL5 (anti-Ku). RBCs of proband's relatives were phenotyped as wild type KEL:-1,2,-3,4,-6,7 (K-,k+,Kp^a-,Kp^b+,Js^a-,Js^b+). The proband and her relatives had the same KEL*2/KEL*2, KEL*4/KEL*4, KEL*7/KEL*7 wild type genotype determined by PCR-RFLP. The G>A mutation on intron 16, position +1 downstream of exon 16 (IVS16+1G>A mutation) was identified by KEL sequencing technique at homozygous state on the proband and at heterozygous state on her sister. The IVS16+1G>A mutation introduce an alternative splicing that may skip exon 16 and create a premature stop codon (TGA) on exon 17. No mutation was detected in the two proband's nieces. Proband's RBCs did not react against all antibodies by flow cytometry: MFI (mean fluorescent intensity) of CD238=3.7, KEL1(K)=7.6, KEL2(k)=7.6, KEL3(Kp^a)=8.5 and KEL4(Kp^b)=8.0 against positive results in control A RBCs [MFI of CD238=77, KEL1(K)=8, KEL2(k)=189.4, KEL3(Kp^a)=9.3 and KEL4(Kp^b)=114.4] and control B RBCs [MFI of CD238=77, KEL1(K)=76, KEL2(k)=50, KEL3(Kp^a)=7.8 and KEL4(Kp^b)=49.6]. Results of the RBC phenotyping by flow cytometry of the proband's sister were CD238=156, KEL2(k)=51 and KEL4(Kp^b)=48. There were no significant differences between results on fresh and frozen samples. The MFI of KEL2 from proband's sister (KEL*2/KEL*2null) RBCs was similar to that observed in control B (KEL*2/KEL*1 heterozygous), 51 and 50, respectively, and lower than the control A (KEL*2/KEL*2 homozygous), 51 and 189.4, respectively, reflecting the single gene dose in the proband's sister and control B, and double gene dose in control A. In conclusion, we report a novel KELnull allele, IVS16+1G>A, detected for the first time in a Brazilian woman with Kell null phenotype. This study was supported by CNPq and FAPESP.