Buried in Jiankui’s unpublished paper were some disturbing pieces of information. In Lulu, only one of the two relevant chromosomes had been properly modified. “We confirmed Nana’s CCR5 gene was edited successfully with frameshift mutations on both alleles and Lulu’s was heterozygous,” he admitted. In other words, Lulu had different gene versions on her two chromosomes, which meant that her system would still produce some of the CCR5 protein.