Amylase Gene Structures in Primates: Retroposon Insertions and Promoter Evolution

Linda C. Samuelson, Ruth S. Phillips, and Lisa J. Swanberg Department of Physiology, University of Michigan

Amylase transcription in the human salivary gland results from the evolutionary juxtaposition of two inserted elements, a y-actin pseudogene and an endogenous retrovirus, to create an unusual salivary-specific promoter. We utilized these structures as molecular tags to characterize the amylase genes in extant primates by polymerase chain reaction amplification of promoter fragments from genomic DNA. Six distinct amylase promoter structures were identified, which allowed us to infer the structures of common ancestors and trace the evolution of the modem human amylase promoters. Our data show that integration of the pseudogene and retrovirus were evolutionarily recent events. The y-actin pseudogene integrated after the divergence of the New World monkeys from the primate ancestral tree, and the retrovirus integrated later, after the divergence of the Old World monkeys. The New World monkey amylase promoter represents the mammalian amylase precursor structure before integration of the two retroposons. Two distinct amylase genes were identified in the Old World monkeys, one with a complete y-actin pseudogene insert and another novel structure with a truncation of the y-actin sequences. We demonstrated abundant amylase expression in the saliva of an Old World monkey, indicating that the endogenous retrovirus is not required for amylase transcription in the primate salivary gland.

Introduction

The human a-amylase gene family provides an ex- cellent model system with which to investigate the evo- lutionary synthesis of a tissue-specific promoter. Molec- ular cloning revealed that distinct gene copies code for pancreatic and salivary amylase (Nishide et al. 1986). The complete gene family, clustered within a 250-kb region on human chromosome 1, contains five active genes, including two pancreatic genes (AMY2A and AMY2B) and three salivary genes (AMYIA, AMYlB, and AMYlC) (Gumucio et al. 1988; Samuelson et al. 1988; Groot et al. 1989). The pancreatic and salivary genes are highly related, with similar intron/exon boundaries and 98% nucleotide sequence identity over their coding regions (Horii et al. 1987). However, the salivary amy- lase genes contain an additional nontranslated exon at their 5’ ends, and the pancreatic and salivary amylase promoters are unrelated. The gene structures are consis- tent with the theory that all five copies arose during evolution from a single ancestral gene through a series of duplications, with subsequent divergence of the pro- moter regions leading to differences in tissue-specific expression (Samuelson et al. 1990).

A striking feature of the human amylase genes is their close association with retroposon inserts. Three distinct promoter structures resulted from the indepen- dent insertion of two transposable elements, a processed y-actin pseudogene and an endogenous retrovirus (Emi et al. 1988; Samuelson et al. 1988, 1990) (fig. 1). Each

Abbreviations: NWM, New World monkey; OWM, Old World monkey; AMY2A, AMY2B, pancreatic amylase genes; PCR, polymerase chain reaction; AMYIA, AMYIB, AMYIC, salivary amylase genes; LTR, retroviral long terminal repeat.

Key words: cY-Amylase, multigene family, promoters, molecular evolution, primates, transposable elements, retroviruses, tissue-specific gene expression.

Address for correspondence and reprints: Linda C. Samuelson, Department of Physiology, University of Michigan, 7761 Med Sci II, Ann Arbor, Michigan 48109-0622, E-mail: lcsam@umich.edu.

Mol. Bid Ed 13(6):767-779. 1996 0 1996 by the Society for Molecular Biology and Evolution. ISSN: 0737-4038

human amylase gene is associated with a y-actin pseu- dogene, situated 0.2 kb upstream of the first coding exon (exon a). This insert has typical features of a processed pseudogene and is 91% identical to the human y-actin cDNA sequence (Samuelson et al. 1990). The AMY2B gene is associated with a complete y-actin pseudogene, approximately 2 kb in length; the other four amylase genes are associated with truncated pseudogene copies, missing the 5 two-thirds of the y-actin sequence. The common pseudogene integration site in all five human amylase genes provides strong evidence that they arose by duplication of a single precursor gene following in- tegration of the y-actin.

The second retroposon identified in the human am- ylase gene cluster is an endogenous retrovirus (Emi et al. 1988; Samuelson et al. 1988). Four copies of the retrovirus are found within the cluster, interrupting the y-actin pseu- dogene upstream of the three AA4YZ genes and the AMY2A gene (fig. 1, filled boxes). The three salivary-specific AMYI genes are associated with complete retroviral ge- nomes, located 0.23 kb upstream from the transcriptional start site and in the opposite orientation to amylase. Re- markably, AMY1 transcription initiates within the y-actin pseudogene, and the salivary amylase promoters are com- posites of pseudogene and retroviral sequences. The pan- creatic AMY2A gene has retroviral sequences in the form of a solo long terminal repeat (LTR). However, transcrip- tion of this gene in the pancreas does not initiate within the pseudogene, but at a site downstream, in a position analogous to the start site of the other pancreatic gene, AMY2B (Samuelson et al. 1988). The AMY2B gene is not associated with a retroviral insert.

The close proximity of the y-actin and retroviral in- serts to the amylase promoters is unusual and suggests that integration of these retroposons may have influenced am- ylase expression. The importance of these elements for the regulation of salivary amylase transcription has been dem- onstrated by Ting et al. (1992) using gene transfer studies in transgenic mice. This study showed that a l.O-kb AMYIC promoter fragment containing sequences derived

767

Table 1

Amylase Promoter Structures in Primates 769

Nucleotide Sequences of Primers Used to Amplify/Sequence Primate Amylase Fragments

Name Primer Nucleotide Sequence (5’ to 3’) Amylase Gene Region” Reference

a2a.. GTGTTCTCCCTTTGTGCAT-ITGGTCAGG a2b. GTGTACACTGCCAGAGGTTTAAAAAGG act TCAGGGTTGGAAAGTCCAAGCCATAGGACC act2 CTGTTATGTGAGAACAATAGGCCCCAGC em.. GGCTATCAATCTATTGCACAGAAG exa2.. TACTGAGCCCAGCAGAACCCAATG gag.. GGTGGAGTGGGCAAGTTGAAAAGACTTGTC mact. CAAATCCCAGAACACTCAGCCCTGACAC 5ga.. GGACTGTGAGCAGATGGCTAGCAATTGC

AMY2A AMY2B AMY1 AMY2 AMYI AMY2 AMY1 AMY2A AMY2B

(S; -1349 to -1322) (S; --2.2 kb) (S; +54 to +83) (S; -277 to -250) (A; --8.5 kb) (A; +67 to +44) (S; -991 to -962) (A; -536 to -563) (S; -914 to -884)

Samuelson et al. 1990 Samuelson et al. 1990 Samuelson et al. 1988 Samuelson et al. 1990 Unpublished datah Gumucio et al. 1988 Ting et al. 1992 Samuelson et al. 1988 Samuelson et al. 1990

Nom-See figure I for a graphic illustration of the positions of the PCR primers m the human amylase promoters. ” AMY1 indicates the three salivary amylase genes (AMYIA, AMYIB, or AMYIC); AMY2 Indicates the two pancreatic amylax genes (AMYZA or AMY2B) S =

sense strand sequence; A = antisense strand sequence. h Nucleotide sequence of the en, region of ERVAIC.

Amplification of Amylase Gene Fragments

Oligonucleotide primers were designed from hu- man amylase sequences (fig. 1, table 1). PCR reactions were carried out in 50 FL containing 100-500 ng of genomic DNA; 10 mM Tris-HCl (pH 8.3); 50 mM KCl, 4 mM MgCI,; 0.01% (w/v) bovine serum albumin; 0.2 mM dATP, dCTP, dGTP, and dTTP (Pharmacia); and 0.2 FM of each primer and Tuq polymerase. The a2alexa2 amplification mix contained 0.5 mM MgCI,. Two dif- ferent thermocycler programs were used. For fragments greater than 1.5 kb, samples were subjected to 35 cycles of 94°C for 25 set, 55°C for 30 set, and 72°C for 1.5 min, with a final extension step at 72°C for 5 min. For fragments less than 1.5 kb, samples were subjected to 35 cycles of 94°C for 20 set, 55°C for 20 set, and 72°C for 20 set, with a final extension step at 72°C for 5 min. After amplification, the PCR products were subjected to agarose gel electrophoresis, stained with ethidium bro- mide, and photographed. In some cases, samples that amplified strongly were diluted prior to agarose gel elec- trophoresis to balance the signals on the gels.

Southern Hybridization

Approximately 1 ng of each PCR product was re- solved by agarose gel electrophoresis before capillary transfer to Zetaprobe membrane (BioRad). Membranes

Table 2 Summary of the Amylase Gene Fragments PCR Amplified from Primate Genomes

Fragment” Primers Primate@ Contents’

AMY-0.6. act/exa2 H, A, OWM amylase, y-actin AMY2B-2.2 a2b/exa2 H, A, OWM amylase, y-actin AMY2B- 1 .O 5ga/exa2 H, A, OWM amylase, y-actin AMYI-1.6.. gag/exa2 H,A amylase, y-actin, LTR AMY l-0.7. a2a/env H, A y-actin, LTR AMYZA- 1.4 a2a/exa2 H, A amylase, y-actin, LTR NWM-0.3 a2b/exa2 NWM amylase OWM-I.0 a2a/exa2 OWM amylase, y-actin

d See Materials and Methods for an explanation of fragment nomenclature. h Fragments were amplified from a minimum of three species within each

group. H = human, A = apes, OWM = Old World monkeys, NWM = New World monkeys.

/ Fragment contents were described by Southern hybridization with AMY (amylase), ACT (y-actin), and LTR probeq.

were hybridized with “‘P-labeled DNA probes specific for amylase (AMY), y-actin (ACT), or retroviral LTR se- quences (fig. 1) at 60°C in hybridization solution (0.5 M NaPO,, 1 mM EDTA, 7% sodium dodecyl sulfate) con- taining denatured salmon sperm DNA (100 pg/mL). Lin- ear DNA fragments used for probes were isolated from a plasmid subclone of cosmid G21 that contains the AMYIC gene (cosmid G21 is described in Gumucio et al. [ 19881). Restriction fragments were purified after electrophoresis in low-melt agarose using the freeze fracture technique de- scribed by Qian and Wilkinson (199 1). The AMY probe was a 0.45-kb EcoRI-Nde I fragment that included exon a and approximately 200 bp of adjacent intronic sequence. The ACT probe was a 0.5-kb Nde I-Hpa I fragment that included sequences from the 3’ end of the y-actin pseu- dogene. The LTR probe was a 0.24-kb Sac I-Bgl I frag- ment that included sequences from the 5’ LTR of the pro- virus. Probes were j2P-labeled using the random primer method of Feinberg and Vogelstein (1983). After hybrid- ization overnight, membranes were washed at 60°C; the final wash solution contained 1 X SSC (0.15 M NaCl and 0.15 M sodium citrate) and 0.1% sodium dodecyl sulfate. Genomic Southern analysis of human and cynomolgus macaque DNAs (10 pg) was performed as described in Samuelson, Isakoff, and Lacourse (1995) using the AMY probe.

Amylase Activity in Saliva Amylase enzyme activity was detected by acrylam-

ide gel electrophoresis and starch-iodine staining using a modification of the procedure of Bloor and Meisler (1980). Saliva or serum samples were diluted in phos- phate buffer (22 mM KH,PO,, 22 mM Na,HPO,, 7 mM NaCl, pH 6.9) to a total volume of 25 FL and amylase isozymes were resolved on a 7.5% acrylamide gel by electrophoresis overnight in buffer containing 12.5 mM Tris and 49 mM glycine (pH 8.1). Gels were subse- quently incubated 30 min at 37°C in a solution of 2.5% starch (Sigma S-2630) in phosphate buffer (w/v), rinsed in water and stained with iodine solution (6 mM potas- sium iodide, 0.8 mM iodine). Amylase enzyme activity was observed as lightly staining areas on the gel.

Sequence Analysis PCR products were isolated from low-melt agarose

gels, purified by centrifugation through Millipore 30,000

772 Samuelson et al.

A. -+I

IPm y -Actin 3’/-pJ HumanAMY2B

7 T a2b eti

*

& e%

Squirrel Monkey

B. y-Actin 5’

AMYZB --2.3 kb TTAAGATT-ITATCAGTAAGAA III1 IIIIIIIII II I

Sq Monkey -215 lTAATAlTlTATCAAGAAACAAlTAATTTCTAAA-GGTCAmAGATTAllTCCATGAAA I IIIIlIIIIIII IIIIIlIIIIII lIIIIIIIlI I

AMYZB A-lTAAll-KTAAAAGGTCAllTAGATGAmCCATGAGA

y-Actin 3'

c. Sq Monkey -1% CAAlTAATT- -TCT-AAAGG TCAll-TAGAT TATTTC-CAT GAAAGACm lTAATAlTCT Hum 2B -197 ..-......- -...A..... . . . . . . . . . . G.....-... ..G....... . . . ..G.... Mus 2.2 -194 ..-...CA.G GC..C.G... G.T..C..T. . . . ..TG.G. ..G..-T..C .A..AG.C.A

Sq Monkey -140 TCACCTGllA AGATTTTTAT TGGTAATCCT TTTCAGATTA TGAATAAACA Gm-CCTC Hum 2B -141 ..T....... G....A.... ..A....... . . . . . . . . . . . ..G...... . . ..GC.... Mus 2.2 -136 . . . . . . . ..C .C..CC.C.C -A.C..AT.. . ..-...... . . ..G...G. . . ..A.T...

Sq Monkey -81 -AAGTATTTA TACATACTAC TATTCACTTT GTAAAATGTG CTTCTTACAG GAATATAAAT

Hum 2B -81 -......... .T...G...A . . ..T..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Mus 2.2 -78 T..A....-. .T....--.A . . ..G....G A........A .C.T.GT... ..-.......

v Sq Monkey -22 AGTTTCTGGA AAGGACACTG ACAAGTKAA GGGAA Hum 2B -22 . . . . . . . . . . . . . . . . . . . . . . ..C..... A.C.. Mus 2.2 -22 G.G.G..AC. G..A...... . . ..C..... A.C..

FIG. 4.-The New World monkey amylase promoter does not contain a y-actin pseudogene. (A) Structural comparison of the NWM-0.3 fragment amplified from the squirrel monkey with the AMY2B-2.2 fragment. (B) Comparison of the nucleotide sequence of the junction regions flanking the y-actin insert with the squirrel monkey promoter, The uninterrupted squirrel monkey sequence is aligned with the 5’ and 3’ junctions of the y-actin pseudogene insert (boxed) associated with AMY2B. (C) Nucleotide sequence alignment of the squirrel monkey amylase promoter with the human AMYZB and mouse An2~2.2 promoters. Nucleotide sequence identities (.) and deleted nucleotides (-) are indicated. The positions of the TATA box (double underlined) and transcription start site of the AMY2B gene in human pancreas (arrowhead) are indicated. Nucleotide numbering for the squirrel monkey gene is based on the human gene numbering. AMY2B sequence is from Gumucio et al. (1988) and Samuelson et al. (1990); Amy2.2 sequence is-from Osborn et al. (1987). _

The amplification of the AMY-O.6 and AMY2B- 2.2 fragments from the apes and Old World monkeys demonstrated that the y-actin pseudogene insertion took place in a common ancestor. Furthermore, the absence of these fragments from the New World monkey sam- ples suggested that the y-actin pseudogene integrated sometime after the divergence of the New World mon- keys. This interpretation was further supported using a third primer pair consisting of another y-actin-specific primer (5ga) paired with exa2 to amplify a l.O-kb frag-

ment from genomes containing an AMY2B-like gene. As observed with the previous two primer pairs, the 5gaI exa2 primer pair amplified a fragment (AMY2B-1 .O) from the human, ape, and Old World monkey DNAs, but not from the New World monkey species (table 2; data not shown). Thus, we have consistently failed to amplify fragments indicative of an amylase-associated y-actin pseudogene from the New World monkeys. This was not due to the inability of the common exa2 primer to align productively with the New World monkey am-

Amylase Promoter Structures in Primates 775

A. LTR IPm y-Actin H a Human AMY2A

7 T

Macaque (OWM- 1 .O)

B. LTR _

AMY2A - 1271 TTGGAAAGAAGTTGTCCTTGAAGC~GTATCTGACATTGTAGCAGGACGAGCCTCAGACAAAACCTCTCA

I I IIIIIlIIIIIlIIIIIIIIIIIIIIIlII II MN --800 TAGAAAAGAAGTTGTCCTTGAAGCTTGTATCTGATATCAGCACTGGACTGTAGAAATGGTTGCTGA~

IIIIIIIIIIIlII IIIIIII I I I I I I I I I I I I AMYZA -804 ATAAAACAAGGAGAGGGTCmGTCTCCTCTCA4TATCAGCACTGGA~GTAGAATGTGTTGCTGA~

1 -

LTR

C.

MM 196

AMY2A 197

AMYZB 197

Sq Monkey 196

Mac 138

AMYZA - 138

AMYZB 138

Sq Monkey - 137

Mac 78 TATTTATTAA

AMYtA - 78 ........ C.

AMYZB - 78 ........ C.

Sq Monkey -78 ....... AC.

Mac 18 AMYZA 18

AMYZB - 18

Sq Monkey - 18

CA-TTAATT- CGAAAAGGTC AmAGATFA TTKCG~ACATTTTA TGGTTfi%ii ..- ..... AT .T .............. ..G. ... ..A ...... ..T .. ..G AT......T. ..- ...... T .T .............. ..G. ... ..A ...... ..T ..... AT ..... ..T .. A ...... T .T- .................... ..A . ..A .. ..T ..... ATA .......

?~Z~+AAGA -~TATGATTGA TAATCC~ CATATTATTA ATAAACAG~ TGCTCTCAAG .C.....G .... ..T ............... ..G.....C. ........... ..C ...... .C.....G .... ..T ............... ..G.....G. G .......... ..C ...... .C ........ ..T.T .. ..G .......... ..G.....G. .......... .-TC ......

TCTGAAAAGG . . . . G . . . . . . . . . G . . . . . . . . . G . . . . .

TGTTAATAl-T TACTl-i-GTAA AATGTGCTTC ‘TTACAAGAAT ATAAATAGTT

. . C . . . . . . .

. . C . . . . . . .

.AC..C....

. . . . . . . . . . . . . . . . . . . . . . . . . G . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . G . . . . . . . . . . . . . .

C . . . . . . . . . . . . . . . . . . . . . . . . G . . . . . . . . . . . . . .

ACACTGACAA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

CT . . . . G.

Frc. 7.-The novel OWM- 1 .O fragment represents an amylase promoter with a truncated y-actin pseudogene and no retroviral sequences. (A) Structural comparison of the OWM-I .O fragment amplified from the macaque with the AMY2A-1.4 fragment. (B) Comparison of the sequence of the junction regions flanking the AMY2A-associated LTR (boxed) with the Japanese macaque promoter. The macaque sequence contains a single copy of the 4-bp region that was duplicated upon insertion of the retrovirus (double underlined). (c) Sequence comparison of the novel macaque promoter with three other primate amylase promoters. Nucleotide sequence identities (.) and deleted nucleotides (-) are indicated. The positions of the TATA box (double underlined), the transcription start site (arrowhead), and the bipartite binding site for the pancreatic transcription factor PTFl (Cockell et al. 1989) (overlined) are based on the human genes. Macaque and squirrel monkey sequences are from this study; AMY2A and AMY2B sequences are from Gumucio et al. (1988).

monkey sample exhibited two distinct bands of activ- Discussion ity, suggesting that they may be products of the two different amylase genes. Since the retrovirus is not In this study we have characterized the amylase associated with the Old World monkey amylase pro- promoters in extant primates to infer the structures of moters, we conclude from this result that the retrovi- amylase gene precursors and predict the pathway of evo- rus is not required for salivary amylase expression. lution of the human salivary amylase promoter. We test-

Amylase Promoter Structures in Primates 777

Salivary Expression

.-H NWM -

I

--o-B OWM +

--_

Y Actin Insertion

Apes + Retroviral Insertion

Human +

FIG. 10.-A summary of the amylase promoter structures detected in this study is diagrammed with the evolutionary relationships of the primate families containing these structures. Promoter structures are labeled with the corresponding human amylase gene: novel promoter structures are unlabeled. The structures detected in the apes (chimpanzee, gorilla, and orangutan) were similar to the previously described human genes. The proposed timing of insertion of the y-actin and retroviral retroposons are indicated on the phylogenetic tree on the right. Salivary amylase expression in each primate group is indicated (+ or -) (Junqueira, Toledo, and Dome 1973; McGeachin and Adin 1982). With the exception of the human genes, the identification of salivary amylase promoters is unknown. The minimal AMYlC gene region (0.7 kb) that confers salivary-specific expression in transgenic mice is indicated by the bar (from Ting et al. 1992). Open boxes, amylase exons; shaded boxes, y-actin pseudogene; filled boxes, endogenous retrovirus.

monkeys (fig. 10). The tinal step in our model is the insertion of the endogenous retrovirus into the truncated y-actin pseudogene to create the AMY1 gene structure observed in the human. This proposed pathway predicts a switch in tissue-specific expression, since the five hu- man gene copies that formed from a single precursor exhibit divergent expression in the pancreas and salivary gland (Samuelson et al. 1988).

Gene expression studies in transgenic mice have demonstrated that salivary-specific transcriptional con- trol elements are located within the amylase-associated provirus, as evidenced by the parotid-specific expression of a transgene containing a 0.7-kb proviral fragment (Ting et al. 1992). This result suggested that salivary specificity is encoded by the retrovirus, and that amylase expression in the salivary gland was determined upon retroviral integration in the primate lineages. However, our demonstration of high levels of amylase activity in the saliva of the macaque contradicts this idea, and in- stead indicates that salivary amylase expression predat- ed the retroviral insertion (fig. 8). Two additional Old World monkey species, the rhesus and mangabey, have also been reported to contain high levels of amylase in their saliva (McGeachin and Adin 1982), indicating that salivary amylase expression is a general property of this primate group. Our analysis showed that neither Old World monkey amylase promoter contained the amy- lase-associated retrovirus (figs. 8 and 10). Thus, an al- ternative mechanism must regulate salivary amylase ex- pression in these primates. Whether the y-actin pseu- dogene insert plays a role in regulating amylase expres- sion is unknown. However, the observation that New World monkeys do not contain the pseudogene or sali-

vary amylase (squirrel monkey and capuchin; Junqueira, Toledo, and Doine 1973; McGeachin and Adin 1982) suggests that the y-actin insertion may have been in- volved in a switch in site of expression from the pan- creas to the salivary gland.

Another example of a switch in tissue specificity among members of a gene family is the human growth hormone cluster, whose five genes are thought to have evolved from a single ancestral gene within the past 10 million years (Barsch, Seeburg, and Gelinas 1983; Chen et al. 1989). This expansion resulted in four genes with a new site of expression in the placenta, and mainte- nance of expression of one growth hormone gene in the pituitary. While the molecular basis for this divergence in tissue specificity is not known, it appears that trans- posable elements, in particular retroviral-like elements, are a potential source of evolutionary alterations in gene expression (Robins and Samuelson 1992). The mouse sex-limited protein gene (sip) (Stavenhagen and Robins 1988), the rat oncomodulin gene (Banville and Boie 1989), and the human histocompatiblity gene HLA- DRB6 (Mayer, Ohuigin, and Klein 1993) show altered gene expression as the result of retroviral-like sequences near the gene promoter. In the example of sip, integra- tion of a C-type retrovirus during rodent evolution in- stalled a hormone-responsive LTR upstream of a dupli- cated complement gene, imposing androgen-regulated expression on the adjacent gene (Stavenhagen and Rob- ins 1988). It would be desirable to confirm possible re- troposon-induced changes in gene expression using a phylogenetic approach, such as the one reported here for the primate amylase genes. Our results indicate that in some instances transcriptional regulatory sequences de-

778 Samuelson et al

tected in modern retroposons may not relate to the mechanism of initial change in gene expression.

All vertebrates express amylase in the pancreas, while expression in the salivary gland is limited to some but not all species of primates, rodents, lagomorphs, and chiropterans (Junqueira et al. 1973). There are two pos- sible models to explain this pattern (Meisler and Gu- mucio, 1986). One model predicts that both salivary and pancreatic amylase existed in the mammalian ancestor with extinction of salivary amylase in several lineages during evolution. Alternatively, the mammalian ancestor might have contained only pancreatic amylase, with sal- ivary expression being acquired independently in certain lineages. Support for independent acquisition of salivary amylase in the rodents and primates is provided by the observation that the human and mouse salivary amylase promoters are unrelated in structure or DNA sequence (Gumucio et al. 1988). The mouse promoter is associ- ated with a nontranslated exon located 7 kb upstream of exon a (Young, Hagenbtichle, and Schibler 1981). In contrast, the human salivary amylase promoter is asso- ciated with a nontranslated exon located 0.5 kb upstream of exon a. Furthermore, the human but not the mouse salivary amylase promoter is the product of the insertion of two transposable elements. Despite the divergent na- ture of the human and mouse salivary amylase promot- ers, human salivary amylase transgenes were expressed in the mouse with the proper tissue specificity (Ting et al. 1992). The nature of the salivary-specific transcrip- tion factors regulating amylase expression in the sali- vary gland has not yet been characterized.

Independent acquisition of salivary amylase in ro- dents and primates would predict that salivary amylase expression can be favored by natural selection. Although the nature of an evolutionary advantage afforded by sal- ivary amylase activity is unclear, it has been hypothe- sized that the agreeable taste of sugars produced by am- ylase enzymatic activity in the mouth could help animals recognize nutritious food sources (Ting et al. 1992). Any increase in digestive efficiency or food recognition could be favored by selection over time.

Acknowledgments

We gratefully acknowledge the contributions of Drs. Slightom, Neiswanger, Tashian, Sampiao, and Goodman for providing primate genomic DNA, and Dr. Joan Keiser for the macaque saliva sample. We thank Priscilla Tucker, Deb Gumucio, and Didi Robins for helpful comments on the manuscript. This research was supported by NIH grant DE09598 and a pilot grant from the University of Michigan Comprehensive Cancer Cen- ter awarded to L.C.S. R.S.P was the recipient of a fel- lowship from the University of Michigan Student Bio- medical Research Program.

LITERATURE CITED

BAILEY, W. J., J. L. SLIGHTOM, and M. GOODMAN. 1992. Re- jection of the “flying primate” hypothesis by phylogenetic evidence from the l -globin gene. Science 256:86-89.

BANVILLE, D., and Y. BOIE. 1989. Retroviral long terminal re- peat is the promoter of the gene encoding the tumor-asso- ciated calcium-binding protein oncomodulin in the rat. J. Mol. Biol. 207:481-490.

BARSCH, G. S., F’. H. SEEBURG, and R. E. GELINAS. 1983. The human growth hormone gene family: structure and evolu- tion of the chromosome locus. Nucleic Acids Res. 11:3939- 3958.

BELL, G. I., J. H. KARAM, and W. J. RUTTER. 1981. Polymor- phic DNA region adjacent to the 5’ end of the human in- sulin gene. Proc. Natl. Acad. Sci. USA 78:5759-5763.

BLOOR, J. H., and M. H. MEISLER. 1980. Additional evidence for the close linkage of amy- and urn?-2 in the mouse. J. Hered. 71:449-45 1.

CHEN, E. Y., Y.-C. LIAO, D. H. SMITH, H. A. BARRERA-SAL- DANA, R. E. GELINAS, and I? H. SEEBURG. 1989. The human growth hormone locus: nucleotide sequence, biology, and evolution. Genomics 4:479-497.

COCKELL, M., B. H. STEVENSON, M. STRUBIN, 0. HAGEN- B~~CHLE, and I? K. WELLAUER. 1989. Identification of a cell- specific DNA-binding activity that interacts with a tran- scriptional activator of genes expressed in the acinar pan- creas. Mol. Cell. Biol. 9:2464-2476.

EMI, M., A. HORII, N. TOMITA, T. NISHIDE, M. OGAWA, T. MORI, and K. MATSUBARA. 1988. Overlapping two genes in human DNA: a salivary amylase gene overlaps with a gamma-actin pseudogene that carries an integrated human endogenous retroviral DNA. Gene 62:229-235.

FEINBERG, A., and B. VOGELSTEIN. 1983. A technique for ra- diolabeling DNA restriction endonuclease fragments to high specific activity. Anal. Biochem. 132:6-13.

GOODMAN, M. 1986. Molecular evidence on the ape subfamily homininae. Pp. 121/014193/ in H. GERSHOWITZ, D. L. RUCKNAGEL, and R. E. TASHIAN, eds. Evolutionary per- spectives and the new genetics. Alan R. Liss, Inc., New York.

GROOT, P C., M. J. BLEEKER, J. C. PRONK, E ARWERT, W. H. MAGER, R. J. PLANTA, A. W. ERIKSSON, and R. R. FRANTS. 1989. The human ru-amylase multigene family consists of haplotypes with variable numbers of genes. Genomics 5: 29-42.

GUMUCIO, D. L., H. HEILSTEDT-WILLIAMSON, T. A. GRAY, S. A. TARLE, D. A. SHELTON, D. A. TAGLE, J. L. SLIGHTOM, M. GOODMAN, and E S. COLLINS. 1992. Phylogenetic foot- printing reveals a nuclear protein which binds to silencer sequences in the human gamma and epsilon globin genes. Mol. Cell. Biol. 12:4919-4929.

GUMUCIO, D. L., D. A. SHELTON, K. BLANCHARD-MCQUATE, T. GRAY, S. TARLE, H. HEILSTEDT-WILLIAMSON, J. L. SLIGHTOM, E COLLINS, and M. GOODMAN. 1994. Differen- tial phylogenetic footprinting as a means to identify base changes responsible for recruitment of the anthropoid gam- ma gene to a fetal expression pattern. J. Biol. Chem. 269: 15371-15380.

GLJMUCIO, D. L., K. WIEBAUER, R. M. CALDWELL, L. C. SAM- UELSON, and M. H. MEISLER. 1988. Concerted evolution of human amylase genes. Mol. Cell. Biol. 8: 1197-1205.

HORII, A., M. EMI, N. TOMITA, T. NISHIDE, M. OGAWA, T. MORI, and M. MATSUBARA. 1987. Primary structure of hu- man pancreatic c*-amylase gene: its comparison with human salivary a-amylase gene. Gene 60:57-64.

JUNQUEIRA, L. C. U., A. M. S. TOLEDO, and A. I. DOINE. 1973. Digestive enzymes in the parotid and submandibular glands of mammals. An. Acad. Bras. Cienc. 45:629-643.

MAYER, W. E., C. OHUIGIN, and J. KLEIN. 1993. Resolution of the HLA-DRB6 puzzle: a case of grafting a de novo-gen-

Amylase Promoter Structures in Primates 779

erated exon on an existing gene. Proc. Natl. Acad. Sci. USA 90: 10720-10724.

MCGEACHIN, R. L., and J. R. ADIN. 1982. Amylase levels in the tissues and body fluids of several primate species. Comp. Biochem. Physiol. 72A:267-269.

MEISLER, M. H., and D. L. GUMUCIO. 1986. Salivary amylase: evolution and tissue-specific expression. Pp. 457-466 in P DESNUELL, H. SJOSTROM, and 0. NOREN, eds. Molecular and cellular basis of digestion. Elsevier Science, Amster- dam.

NISHIDE, T., Y. NAKAMURA, M. EMI, T YAMAMOTO, M. OGA- WA, T. MORI, and K. MATSUBARA. 1986. Corrected se- quences of cDNAs for human salivary and pancreatic (Y- amylases. Gene SO:37 l-372.

OSBORN, L., M. P ROSENBERG, S. A. KELLER, and M. H. MEIS- I.ER. 1987. Tissue specific and insulin-dependent expression of a pancreatic amylase gene in transgenic mice. Mol. Cell. Biol. 7:326-334.

QIAN, L., and M. WILKINSON. 199 1. DNA fragment purifica- tion: removal of agarose 10 minutes after electrophoresis. Biotechniques 10:738.

ROBINS, D. M., and L. C. SAMUELSON. 1992. Retroposons and the evolution of mammalian gene expression. Genetica 86: 191-201.

SAMUELSON, L. C., M. S. ISAKOFF, and K. A. LACOURSE. 1995. Localization of the murine cholecystokinin A and B recep- tor genes. Mamm. Genome 6:242-246.

SAMUELSON, L. C., K. WIEBAUER, D. L. GUMUCIO, and M. H. MEISLER. 1988. Expression of the human amylase genes: recent origin of a salivary amylase promoter from an actin pseudogene. Nucleic Acids Res. 16:8261-8275.

SAMUELSON, L. C., K. WIEBAUER, C. M. SNOW, and M. H. MEISLER. 1990. Retroviral and pseudogene insertion sites reveal the lineage of human salivary and pancreatic amylase genes from a single gene during primate evolution. Mol. Cell. Biol. lo:25 13-2520.

STAVENHAGEN, J. B., and D. M. ROBINS. 1988. An ancient provirus has imposed androgen regulation on the adjacent mouse sex-limited protein gene. Cell 55:247-254.

TING, C.-N., M. I? ROSENBERG, C. M. SNOW, L. C. SAMUEL- SON, and M. H. MEISLER. 1992. Parotid-specific expression of a human salivary amylase gene in transgenic mice. Genes Dev. 6:1457-1465.

YOUNG, R. A., 0. HAGENB~CHLE, and U. SCHIBLER. 1981. A single mouse cw-amylase gene specifies two different tissue- specific mRNAs. Cell 23:45 I-458.

THOMAS EICKBUSH, reviewing editor

Accepted February 29, 1996