Contractions and expansions in gene size and number - Human Gene Evolution - page 341

section 9.5).

Alu sequence polymorphisms. Once inserted, specific Alu sequences have often
been relatively stable in terms of their location during primate evolution (Sawada
et al. , 1985). This notwithstanding, some human Alu sequences are polymorphic
in terms of their presence or absence (Batzer et al. , 1994; Edwards and Gibbs,
1992; Kass et al. , 1994; Meagher et al. , 1996; Milewicz et al. , 1996; Tishkoff et al. ,
1996; Zucman-Rossi et al. , 1997), a situation which is sometimes also found in
other primates (Bailey and Shen, 1993). Some of these polymorphisms may be of
functional significance e.g. a common insertion/deletion polymorphism
(0.40/0.60) within intron 16 of the human angiotensin I converting enzyme
( DCP1 ; 17q23) gene, explicable in terms of the presence or absence of a 287 bp
Alu repeat, is known to have an important influence on serum enzyme concentra-
tion (Rigat et al. , 1990). Alu sequence retrotransposition is also an occasional cause
of genetic disease (e.g. Muratani et al. , 1991; Vidaud et al ., 1993; Wallace et al. ,
1991).

Alu sequence target sites. Although Alu sequences occur on average every 3-6
kb, there are several examples of regions that appear to be preferential target sites
for Alu sequence insertion in mammalian genes. Thus, a 40 kb region, spanning
the spermatid-specific protamine genes PRM1, PRM2 and the transition protein
( TNP2 ) gene (16p13.2), contains a total of 42 Alu sequences (Nelson and Krawetz,
1994). Similarly, a 22 kb region telomeric to the HLA-B-associated transcript 2
(BAT2; D6S51E ; 6p21.3) gene in the HLA class III locus contains 42 Alu repeats
(Iris et al. , 1993), whilst a 2.2 kb segment 5 to the human lysozyme ( LY Z ; 12) gene

contains four such repeats (Riccio and Rossolini, 1993).

Alu sequences within protein-coding sequences. Many Alu sequences are
found within introns and therefore this repeat is represented in heterogeneous
nuclear RNA. Alu sequences are however also found at different locations within
mRNA-homologous sequences, the majority occurring within the untranslated
regions (UTRs). Thus, Yulug et al. (1995) reported that 5% of full-length human
cDNAs contained an Alu sequence, with 82% and 14% of these being located in
the 3 UTR, respectively. In a few cases, however, Alu sequences have

been incorporated into the coding sequences of human genes and have therefore
altered the amino acid sequences of the encoded proteins. Thus, a 279 bp Alu
sequence spans 103 bp of the coding region of a zinc finger protein ( ZNF91 ;
19p12) gene and extends 166 bp into the 3

UTR and 5

UTR (Yulug et al. , 1995). Similarly, 110
bp of Alu sequence lies within the coding a region of the lectin-like type II inte-
gral membrane protein ( KLRC1 ; 12) gene with 43 bp extending into the 3 UTR

(Yulug et al. , 1995). An Alu sequence is entirely contained within the coding
region of the protein serine/threonine kinase stk2 ( STK2 ; 3p21.1) (279 bp Alu ;
Yulug et al. , 1995) gene. Finally, and perhaps most dramatically, two Alu
sequences (both with poly(A) tails) are entirely contained within the coding
region of the regulator of mitotic spindle assembly 1 ( RMSA1 ; 17p11.2-p12) gene
accounting for 111 amino acids of its coding potential, some 40% of the total
(Margalit et al. , 1994).