The distal 30 nt of LILRB1 exon 1 accounts for the poor protein expression by LILRB1 exon 1–containing constructs. (A) Schematic of constructs used to isolate the region within LILRB1 exon 1 responsible for poor protein expression. (B-D) Constructs used for transfections contained the LILRB1 coding region preceded by the portions of the LILRB1 5′-UTR indicated here. Sites in exon 1 that, in some constructs (B,D), were mutated are indicated by an asterisk (*) for ATG sequences and a double-dagger (‡) for the ARE sequence. (B) Graph comparing LILRB1 5′-UTR plus coding region constructs. Mutations were introduced into an LILRB1 5′-UTR construct containing the full-length exon 1. For mut ATG1, ATG2, ATG3, and ATG123, the 3 potential start codons found in LILRB1 exon 1 were changed to AAG. For mut ARE, the ATTTA sequence was changed to ATCTA; n = 4 transfections. (C) Graph comparing LILRB1 5′UTR plus coding region constructs. Progressively truncated LILRB1 exon 1 constructs were compared with an LILRB1 coding region construct; n = 3 transfections. (D) Graph comparing LILRB1 5′-UTR plus coding region constructs. The ATG and ARE sites in the ex1Δ145 construct were mutated, and CD85j expression was compared with the unmutated construct and an LILRB1 coding region construct; n = 5 transfections.