HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Full Text (PDF) All Versions of this Article: 18/11/1288 03-1703fjev1    most recent Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by MALONEY, B. Articles by LAHIRI, D. K. Search for Related Content PubMed PubMed Citation Articles by MALONEY, B. Articles by LAHIRI, D. K.

Presence of a "CAGA box" in the APP gene unique to amyloid plaque-forming species and absent in all APLP-1/2 genes: implications in Alzheimer’s disease

BRYAN MALONEY, YUAN-WEN GE, NIGEL GREIG and DEBOMOY K. LAHIRI^,,1

Departments of
Psychiatry and of
Medical and Molecular Genetics, Institute of Psychiatric Research, Indiana University School of Medicine, Indianapolis, Indiana, USA; and
Laboratory of Neurosciences, National Institute on Aging, NIH, Baltimore, Maryland, USA

¹Correspondence: Indiana University School of Medicine, Institute of Psychiatric Research, 791 Union Dr., Indianapolis, IN 46202, USA. E-mail: dlahiri{at}iupui.edu

SPECIFIC AIMS

The amyloid-ß precursor protein (APP) is the source of toxic amyloid-ß peptide (Aß), which is associated with Alzheimer’s disease (AD). A "CAGA" sequence within the 5'-untranslated region (5'-UTR) of the APP gene is predicted to be the loop of a stem loop structure in a portion of the 5'-UTR that has been implicated to participate in gene regulation. We investigate whether this CAGA box is unique to the APP genes of humans and other species that naturally form pathologic amyloid brain plaque or if it is common to all mammals or even to all species’ APP and homologues with no pathological associations. If the former is the case, what evidence of this site’s activity can be found?

PRINCIPAL FINDINGS

1. Detection and alignment of 5'-UTRs of APP and APP-like sequences in 27 different species
We compared 27 different DNA sequences from the GenBank database. These consisted of 1) twelve APP sequences from 12 different species; 2) five APLP1/APLP2 sequences from three species; 3) two invertebrate APP-like protein sequences; and 4) six prion (PrP) sequences from six species. We included two prion sequences from yeast (see full text online for GenBank accession numbers). Within mammalian APP sequences, human 5'-UTR has 97% homology with monkey, 89% with guinea pig, 84% with rat, and 82% with mouse. Similarity falls with nonmammalian vertebrate APP sequences, and becomes very low when human APP (HsAPP) is compared with APLP and invertebrate APP-like sequences. Mammalian APP has closer similarity to APLP1 and APLP2 than it does to nonmammalian APP superfamily members. APLP1 and APLP2 have greater similarity within APLP1/APLP2 than to their same-species APP. Similarity between the human APP sequences and prions is equal to that between human APP and invertebrate APP-like sequences.

2. Alignment of 5'-UTR sequences and identification of the "amyloid" CAGA in specific species
APP superfamily CAGA box correspondences from Multalin alignment are displayed in Fig. 1 A. Three different CAGA boxes had some level of cross-species representation. The CAGA box marked "mammal" (nucleotides 59–62 in HsAPP) is shared by all mammalian APP sequences surveyed. CAGA box "murinae" (16–19 in mouse sequence) appeared only in rat and mouse APP sequences. Note the CAGA box at 83–86 in HsAPP, which is the stem loop site mentioned before. It occurs only in amyloid plaque-forming species (human, monkey, and guinea pig) and has been named amyloid for that reason. Extending this comparison to mammalian non-APP proteins within the superfamily shows that all APLP1 and APLP2 lack CAGA box amyloid, although HsAPLP2 has a GAGA sequence in a similar location, and mouse APLP2 (MAPLP2) and rat APLP2 (RnAPLP2) have both a GAGA sequence and a CAGA box immediately upstream of this position. APLP1 sequences lacked all types of CAGA boxes.

View larger version (43K): [in this window] [in a new window]   Figure 1. APP superfamily 5'-UTR alignment and HsAPP vs. PrP alignment. A) APP superfamily 5'-UTR sequences were aligned with Multalin and the alignment was used to generate a graphic depiction of homologous CAGA boxes. CAGA boxes with a full 4 nt match are in solid color; those with a single nt mismatch are diagonally hatched. Sequences begin at RNA TSS or first nt in published mRNA and end immediately upstream of the translation start. Selected positions in the human sequence are indicated. Vertical bars indicate 10 nt sequence length. Regions corresponding to the human IL-1-responsive element are outlined in solid black for mammalian APP sequences. Selected putative SP1 binding sequences are enclosed in a broken outline. Positions of the murinae, mammal, and amyloid CAGA boxes are indicated. B) Eight prion sequences were aligned with the APP superfamily alignment using Multalin. The alignment was then used to generate a graphic depiction of homologous CAGA boxes.

3. Flanking sequence around amyloid CAGA: presence of iron-responsive element (IRE), interleukin–1 (IL-1) -responsive, and putative SP1 binding and sites
Putative SP1 binding sites appear at 79–84 and 127–136 in the HsAPP sequence. These sites have homologues in APP sequences from aligned amyloid plaque-producing species but none in APP from non-plaque-producing species. An IL-1 acute box-responsive element appears immediately downstream of CAGA box amyloid, and a bipartite (IRE) includes CAGA box amyloid. Within plaque-forming species, all primate sequences are identical. The guinea pig sequence has two single nucleotide (nt) differences, both near the 5' end of the site. Rat and mouse sequences share a single C-to-T difference from the human sequence, and the rat sequence has two additional differences from the human. The region appears to be unique to mammalian APP members of the superfamily.

4. CAGA sequences in prion and APP 5'-UTR
Prion (PrP) sequences showed no direct CAGA box correspondence with APP superfamily members; yeast sequences were even more distant (Fig. 1B ). A CAGA box does occur in the HsPrP sequence in close proximity to the PrP start codon.

5. APP 5'-UTR fragments with the amyloid CAGA box bind to nuclear proteins
Electrophoretic mobility shift assay (EMSA) with nuclear extracts from rat pheocytochroma (PC12) cells, using a region of the human APP 5'-UTR that contained the amyloid CAGA box (54/144) (transcription start site (TSS) as +1) as well as a fragment lacking the amyloid CAGA box (100/144) (Fig. 2 A), showed a sharp DNA–protein band with the 54/144 fragment (solid arrow) but no corresponding DNA–protein interaction with the smaller fragment (Fig. 2B ). A much weaker interaction was present for both fragments (dashed arrow).

View larger version (50K): [in this window] [in a new window]   Figure 2. Functional analysis of amyloid CAGA box in human APP 5'-UTR. A) Schematic of fragments used in assays. A SmaI/NruI fragment of the human APP 5'-UTR that contains CAGA box amyloid and a BamHI/NruI fragment of the human APP 5'-UTR lacking CAGA box amyloid were purified for EMSA or cloned into plasmid pSV2CAT for reporter gene assay. B) EMSA results. Both fragments were radiolabeled, incubated with PC12 nuclear extracts and analyzed on native 5% polyacrylamide gel. Lanes 1, 2 are 5'-UTR fragments without nuclear extracts. In addition to the 54/144 fragment, lane 1 has a 67 nt unrelated fragment to act as size marker. Lanes 3, 4 were incubated with PC12 extracts. The 54/144 fragment produced distinct DNA–protein interaction (solid arrow); both fragments had nonspecific DNA–protein interaction (dashed arrow). C) CAT functional assay. pSV2CAT backbone constructs with either the 54/144 or 100/144 fragments were transfected into PC12 cells. Cell extracts were assayed for CAT by thin-layer chromatography followed by autoradiography. Lane 1: pSV2CAT-, lane 2: 54/144 SV2CAT-, lane 3: 100/144 SV2CAT-transfected cell extract. Solid arrows indicate acetylated chloramphenicol derivatives and the broken arrow unacetylated chloramphenicol. D) CAT assay was normalized against cotransfected ß-galactosidase activity for each extract. Results were analyzed with the SAS statistical package, using the Waller-Duncan k-ratio test. Samples in same category are not significantly different from each other.

6. APP 5'-UTR activity is localized to the region that contains the amyloid CAGA box
We constructed fusion plasmids containing the APP 5'-UTR plus a reporter gene from pSV₂CAT, which has the chloramphenicol acetyltransferase gene (CAT) driven by the SV40 promoter. This was modified by adding the 54/144 or the 100/144 5'-UTR regions. These constructs were transfected into PC12 (rat adrenal) cells. Assay for CAT activity revealed loss of expression when the 54/99 region containing CAGA box amyloid was deleted (Fig. 2C, D ). Alternatively, the loss of a band in EMSA (Fig. 2B ) and reduction of expression (Fig. 2C ) may be due to loss of several elements other than the CAGA box in the 54/100 region (Fig. 3 ).

View larger version (21K): [in this window] [in a new window]   Figure 3. Model of CAGA box-mediated APP gene regulation. Numbers shown are nt, TSS being +1. Solid black represents a CAGA sequence (e.g., 83/86, panel A). Falling diagonal represents an "AGAC" sequence (e.g., 16/19 in 3B) CAGA box. Square crosshatch is a "TCTG" CAGA box (e.g., –31/–28, A). Diagonal crosshatch represents putative SP1 binding sites (e.g., 47–54, A). Vertical bars represent an IL-1-responsive site (e.g., 86–104, A). The APP TGF-ß-responsive element (which overlaps both the mammal CAGA box and one putative SP1 site in a multiple site sequence) is indicated with a dotted box. Check pattern represents the APBß region (–94/–83, A). Diagonal check represents a potential stem loop structure upstream of the TSS (–28/–3, A) IRE sequences are indicated by brackets. A) Model of TGF-ß mediation of human APP 5'-UTR regulation. TGF-ß binding to its receptor results in phosphorylation of Smads2/3 in complex, phosphorylated Smad3, then recruits Smad4, resulting in Smad3/4 complex formation and transport into the nucleus, where it stimulates CTCF binding at the APBß site (solid arrows). Additional and repeated TGF-ß stimulation may also result in sufficient Smad3/4 complex formation that binding to 5'-UTR CAGA boxes significantly recruits SP1, which stimulates greater APP expression (dashed arrows). B) Schematic of mouse APP 5'-UTR. Potentially disadvantageous position of mouse and murinae CAGA boxes binding with Smads3/4 may present steric hindrance to other factors binding near the transcription start site and/or relative positions of CAGA boxes to SP1 sites is less efficient than for the human sequence. Expression is more difficult to up-regulate and more likely to remain at basal level. C) Human PrP schematic. CAGA boxes suggest a role for Smad-mediated regulation. The sites have no comparable potential IL-1-responsive element or APPtre, although they are much closer to the start codon than is the APP gene cassette.

CONCLUSIONS AND SIGNIFICANCE

Overexpression of the APP gene has been implicated in AD and Down’s syndrome. Herein we report that APP overexpression may be additionally mediated through its 5'-UTR. This provides a mechanism of excess APP and Aß production under inflammatory conditions, leading to amyloid deposition, and suggests a possible path to explore for therapeutic intervention.

The APP mRNA 5'-UTR has been shown to respond to transforming growth factor-ß1 (TGF-ß) stimulation in the absence of sequences upstream of the TSS. TGF-ß signaling is through the Smad pathway, which requires a CAGA box. Here we report a particular CAGA sequence that is unique to the four genes of 19 APP superfamily members that form Aß plaque. Within the 5'-UTR of the human APP gene, elimination of a fragment that contains the CAGA sequence of interest also eliminates protein–DNA interaction in EMSA and reduces expression in a reporter gene system to levels indistinguishable from the unmodified pSV2CAT vector.

We selected the APP 5'-UTR to study as it shows novel gene functional elements: 1) an IL-1-responsive element that confers translational control of APP protein synthesis, 2) an IRE sequence similar to one found in the 5'-UTR of ferritin, 3) direct stimulation by TGF-ß of reporter gene expression in a fusion clone of the 54/144 region of the 5'-UTR, and 4) promoter activity of the first 104 nt of the 5'-UTR.

Based on our work and that of others, we propose a dual regulatory role for the 5'-UTR’s sequence as 1) a transcriptional control point, i.e., TGF-ß-responsive element (APPtre), and 2) a post-transcriptional IL-1 element in the cytoplasm, one or both mediated by the linchpin amyloid Smad3/4 binding site in plaque-forming APP species (Fig. 3) . In addition, we suggest that the 5'-UTR stem loop CAGA box of HsAPP is worthy of further study as a potentially more widespread regulatory model.

FOOTNOTES

To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.03-1703fje; doi: 10.1096/fj.03-1703fje

This article has been cited by other articles:

				L. A. Shakes, T. L. Malcolm, K. L. Allen, S. De, K. R. Harewood, and P. K. Chatterjee Context dependent function of APPb enhancer identified using enhancer trap-containing BACs as transgenes in zebrafish Nucleic Acids Res., November 1, 2008; 36(19): 6237 - 6248. [Abstract] [Full Text] [PDF]

				A. Marutle, M. Ohmitsu, M. Nilbratt, N. H. Greig, A. Nordberg, and K. Sugaya Modulation of human neural stem cell differentiation in Alzheimer (APP23) transgenic mice by phenserine PNAS, July 24, 2007; 104(30): 12506 - 12511. [Abstract] [Full Text] [PDF]

				D. K. Lahiri, D. Chen, B. Maloney, H. W. Holloway, Q.-s. Yu, T. Utsuki, T. Giordano, K. Sambamurti, and N. H. Greig The Experimental Alzheimer's Disease Drug Posiphen [(+)-Phenserine] Lowers Amyloid-beta Peptide Levels in Cell Culture and Mice J. Pharmacol. Exp. Ther., January 1, 2007; 320(1): 386 - 396. [Abstract] [Full Text] [PDF]

				S. Bandyopadhyay, J. Ni, A. Ruggiero, K. Walshe, M. S. Rogers, N. Chattopadhyay, M. A. Glicksman, and J. T. Rogers A High-Throughput Drug Screen Targeted to the 5'Untranslated Region of Alzheimer Amyloid Precursor Protein mRNA J Biomol Screen, August 1, 2006; 11(5): 469 - 480. [Abstract] [PDF]

This Article Full Text (PDF) All Versions of this Article: 18/11/1288 03-1703fjev1    most recent Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by MALONEY, B. Articles by LAHIRI, D. K. Search for Related Content PubMed PubMed Citation Articles by MALONEY, B. Articles by LAHIRI, D. K.