|Home | About | Journals | Submit | Contact Us | Français|
Apolipoprotein B mRNA is edited at cytidine 6666 in the enterocytes lining the small intestine of all mammals; converting a CAA codon to a UAA stop codon. The conversion is ~80% efficient in this tissue and leads to the expression of the truncated protein, ApoB48, essential for secretion of dietary lipid as chylomicrons. Caco-2 cell raft cultures have been used as an in vitro model for the induction of editing activity during human small intestinal cell differentiation. This induction of apoB mRNA editing has been ascribed to the expression of APOBEC-1. In agreement our data demonstrated differentiation-dependent induction of expression of the editing enzyme APOBEC-1 and in addition we show alternative splicing of the essential auxiliary factor ACF. However transfection of these editing factors in undifferentiated proliferating Caco-2 cells was not sufficient to induce robust apoB mRNA editing activity. Only differentiation of Caco-2 cells could induce more physiological like levels of apoB mRNA editing. The data suggested that additional regulatory mechanism(s) were induced by differentiation that controlled the functional activity of editing factors.
Apolipoprotein B (ApoB) is a non-exchangeable structural component of intestinally derived chylomicrons and of liver derived very low density lipoprotein particles (VLDL). The truncated variant of ApoB lipoprotein (ApoB48) is expressed via post-transcriptional RNA editing [1; 2] which involves a site-specific deamination of cytidine (C6666) to form uridine , thereby creating an in-frame translation stop codon, UAA, from a glutamine codon, CAA [1; 2]. The mammalian small intestine constitutively edits ≥85% of apoB mRNA and stores ApoB48 for the assembly and secretion of chylomicrons containing dietary lipids [1; 2; 4; 5].
Site-specific C to U apoB mRNA editing requires a macromolecular complex assembly of proteins (editosome) that minimally contains a catalytic subunit, the cytidine deaminase APOBEC-1 (27 kDa) , and an essential RNA binding protein known as APOBEC-1 Complementation Factor (ACF64) . Insulin-stimulated alternatively splicing of acf mRNA  gives rise to a 65 kDa protein (APOBEC-1 Stimulating Protein, ASP ). ACF64 and ACF65 bind to apoB mRNA and APOBEC-1 and have equivalent ability to support apoB mRNA editing in vitro [8; 10; 11].
Developmental induction of apoB mRNA editing occurs in the fetal intestine beginning with gestation day 17–18 or week 11, in rat or human intestine respectively, and reaches adult levels in rat by gestation day 21–22 [12; 13; 14; 15; 16]. An interesting observation relative to the findings in the current study is that the rate of apobec-1 mRNA expression lags behind the rate of induction of editing, suggesting that editing activity is activated as soon as APOBEC-1 has been expressed. The findings suggest that apobec-1 gene expression is an important if not the dominant element determining the onset of intestinal editing activity.
In the present study we evaluated whether APOBEC-1 expression was the sole determinant for the induction of editing activity during intestinal cell development. We show that apoB mRNA editing could not be detected until apobec-1 mRNA expression was observed in differentiating Caco-2 cells, beginning approximately seven days post plating in differentiating conditions. At this time a shift in expression of alternatively spliced auxiliary protein mRNAs encoding ACF65 and ACF64 was also initiated. The data supported the hypothesis that APOBEC-1 expression and auxiliary protein alternative splicing were regulated during differentiation and responsible for the onset of editing activity. However, transient expression of APOBEC-1 alone or together with ACF64 or ACF65 in undifferentiated and proliferating cells was not sufficient to induce significant changes in apoB editing activity. These results suggested that newly expressed APOBEC-1, ACF64 and ACF65 are not fully functional until other activating regulatory mechanisms have been evoked during the course of Caco-2 differentiation.
Caco-2 cells were purchased from ATCC (Manassas, VA) and cultured as recommended (EMEM w/ 2mM L-glutamine (Gibco, Grand Island, NY), 20% FBS (Gibco), 1 mM Sodium Pyruvate, 0.1 mM Non-essential amino acids (Gibco)). To induce differentiation proliferating Caco-2 cells were plated on 1 µM PET filter inserts (BD Bioscience, San Jose, CA) at 5 × 104 cells per filter with differential media bathing the apical and basal surfaces. Standard media conditions were maintained on the basal surface and serum free media bathed the apical surface, as previously described . Transient transfection of proliferating Caco-2 cells with pcDNA3 encoding HA epitope tagged APOBEC-1 and V5 tagged ACF64 or 65, previously described , were conducted using FuGene 6 (Roche, Indianapolis, IN) per manufacture protocol.
Total RNA was isolated from Caco-2 cells at one week intervals up to 21 days following plating on filter inserts using TRI-REAGENT (MRC, Cincinnati, OH) per manufacture protocol. ApoB mRNA editing levels were assessed on 25 ng of cDNA created by oligo dT primed RT-PCR using the primers MS2, CTACTTCCACTTTTGTTAAAATC and MS3, GAAAATACAGAGCAGCCCCTG, in a poisoned primer extension assay as previously described . Transcript levels were observed by RT-PCR. 2 µg of total RNA isolated during the differentiation time course were primed using oligo dT for reverse transcription then amplified with primers CTGGGAGTTTGACGTCTTC and CCAGCAGTGATAATACTCTG or TCTCTCTTTCTGGCCTGGAG and CTATCTTGGGCTGTGACAAAG to amplify human apobec-1 or β-2-microglobulin, respectively. Products were observed by incorporation of a 32P end labeled 3’ primer in the PCR reaction at a ratio of 1:10 with unlabeled primer using the cycling conditions; 94° C, 3 min., 94° C, 30 sec, 53° C, 60 sec., 72° C, 90 sec for 30 cycles. The reactions were resolved on agarose gels, dried and quantified by Phosphorimager scanning densitometry. Transcript levels of acf64 and acf65 were assessed in the same manner using the primers CTTTGGGCCAAGTTTATGATCC and CTGTGTATGCCAAATAGCCACG, as previously described  and resolved on a 2% agarose Metaphor gel (BMA, Rockland, ME). Immunological Techniques: Caco-2 cells were harvested using the NE-PER kit (PIERCE, Rockford, IL) in the presence of a protease inhibitor cocktail (USB, Cleveland, OH) and then were pooled. Pooled extracts were then placed into two separate immunoprecipitation reactions with a polyclonal anti-ACF antibody recognizing the epitope HTLQTLGIPTEGGD common to the C-terminus of both ACF64 and 65 (ASP) (Bethyl Laboratories, Montgomery, TX). Elutants were resolved by SDS-PAGE and blots were probed with either a polyclonal anti-ACF N-terminus or ACF65 specific antibodies (Bethyl Laboratories) recognizing the epitopes NHKSGDGLSGTQKE or CEIYMNVPVGA, respectively. Equal aliquots of IP material were resolved by SDS-PAGE and probed with anti-β-actin (Sigma, St Louis, MO) for normalization of loading. Detection of APOBEC-1 and ACF64 or 65 in transient transfections utilized anti-HA (Covance, Princeton, NJ) and anti-V5 (Invitrogen, Carlsbad, CA), respectively.
Caco-2 cells are a model system for human intestinal cell (enterocyte) differentiation [14; 15; 16]. Monolayers of polarized columnar enterocytes form over 14 to 21 days in culture when grown on filter membranes (raft cultures) exposed to rich media on the basal surface and minimal media on the apical (see Methods). Differentiated Caco-2 cells secrete chylomicron-like lipoprotein particles assembled on apolipoprotein B48 from their basal membranes exposed to the rich media. This differentiation phenotype results from site-specific C to U editing of apoB mRNA giving rise to a translation stop codon that limits expression of the ApoB protein to 48% of the length of that encoded by the apoB gene [1; 2; 19]. Induction of this phenotype requires the expression of both APOBEC-1, the cytidine deaminase that catalyzes editing  and ACF, the auxiliary protein (APOBEC1 Complementation Factor) [9; 20] that is required for RNA editing site recognition [12; 14; 16; 21; 22].
As anticipated, proliferating Caco-2 cells expressed apoB mRNA but did not demonstrate significant levels of editing (Figure 1A and Table 1). Upon differentiation, apoB mRNA editing increased up to 20-fold. Semi-quantitative RT-PCR (see Methods) for apobec-1 mRNA normalized to β-2-microglobulin mRNA (previously demonstrated to be unaffected by differentiation  and used to normalize cell input in RT-PCR reactions) demonstrated very low apobec-1 mRNA expression until 14 days in culture; after which apobec-1 mRNA expression was markedly elevated (Figure 1B). APOBEC-1 protein levels were not determined as its expression level was below the detection limits of available antibodies. However, a comparison of the induction of apoB mRNA editing activity (Figure 1A) with apobec-1 mRNA expression (Figure 1B) suggested that the expression of the deaminase was rate limiting and corroborating an early report  suggesting that it was a key differentiation-dependent determinant for the induction of editing activity.
In contrast to APOBEC-1, there was significant expression of ACF in proliferating Caco-2 cells and the expression of ACF was maintained throughout differentiation as evident from both the semi-quantitative RT-PCR and western blotting with ACF peptide-specific polyclonal antibodies (see Methods) (Figure 2A and 2B, respectively). Several ACF alternatively spliced mRNAs have been identified and the ability of their translation products to complement APOBEC-1 has been determined [8; 11; 23; 24]. Among the variants, ACF64 and ACF65 exhibit the most robust complementation of APOBEC-1 [8; 9; 11; 20]. ACF65 has an additional eight amino acids not found in ACF64 that enable acf65 mRNA to be differentiated from acf64 mRNA by radiolabeled RT-PCR (Figure 2A). The data suggest that the abundance of both mRNAs in proliferating Caco-2 cells was approximately equivalent (Figure 2A). Following 21 days of differentiation the acf64 mRNA was the prominent spliced variant (acf64:acf65 ratio of 1.8:1) consistent with the finding that acf64 mRNA is the predominant spliced variant in fetal and adult human intestine  as well as in rat liver, rat intestine and HepG2 human hepatoma cells .
Western blotting with anti-ACF65 antibodies (a peptide-specific polyclonal antibody reactive with the eight amino acids unique to ACF65) and with an ACF64/ACF65 common antibody (a peptide-specific polyclonal antibody reactive with the N-terminus common to ACF64 and ACF65) demonstrated that both variants were expressed in Caco-2 cells (Figure 2B upper and lower panels, respectively). There was reduction in ACF65-specific western signal relative to the signal from β-actin (used as a cell number loading control). A corresponding loss of ACF64/ACF65 N-terminus western signal was not observed, which suggested that an increasing amount of this signal came from ACF64 over the time course of differentiation while the signal from ACF65 decreased. Taken together, the data suggested that enhanced APOBEC-1 expression and possibly changes in the relative abundance of ACF65 and ACF64 were integral to the induction of apoB mRNA editing activity during Caco-2 cell differentiation.
To directly test this hypothesis under defined in vitro conditions, apoB mRNA editing activity was quantified in undifferentiated and proliferating Caco-2 cells following transfection and ectopic expression of HA-tagged APOBEC-1. Though APOBEC-1 expression was apparent by western blotting (Figure 3A), little change occurred in the ability of proliferating cells to support apoB mRNA editing activity (Figure 3B and Table 1). These data suggested editing factors where not assembling as functional editosomes even thought transfected cells had abundant ectopic APOBEC-1 and endogenous ACF.
Co-transfection of undifferentiated and proliferating Caco-2 cells with apobec1 and either acf64 or acf65 cDNAs lead to the expression of the encoded proteins (Figure 3A). Despite abundant amounts of APOBEC-1 and ACF64 or ACF65, apoB mRNA editing activity was not stimulated (Figure 4B and Table 1). This surprising result suggested that although there were measurable changes in APOBEC-1 and ACF64/ACF65 abundance over the course of differentiation, the overexpression of these proteins in proliferating cells must not have been sufficient to induce a functional interactions of editing factors with apoB mRNA and RNA editing activity. We hypothesize that over expression of editing factors was not sufficient to induce editing activity in proliferating cells because: (1) they lacked the expression of yet to be identified differentiation-specific factor(s), (2) they could not carry out the necessary post-translational modifications and/or (3) an inhibitory factor was present. These findings augment the current model for the induction of editing during enterocyte differentiation in which the expression of editing factors is the primary, if not the sole determinant for the onset of activity.
It is of interest in this regard to consider potential parallels in the regulatory mechanisms that have been observed during liver development. Hepatic apoB mRNA editing did not demonstrate significant induction until 15 days postpartum and only reached adult levels by 25 days postpartum [12; 22; 25]. During liver development the induction of apobec-1 mRNA expression preceded the induction of editing. The absence of a coincident increase in apoB mRNA editing upon the induction of rat hepatic apobec-1 mRNA suggested that the enzyme was not translated or had been maintained in an inactive state by an inhibitory activity and/or that appropriate auxiliary factors needed to be expressed before editosomes could be assembled.
ApoB mRNA editing activity is considered to be constitutive in the adult intestine whereas it is known to be highly regulated in adult rodent liver. Hepatic APOBEC-1 expression appeared to account for most of the observed changes in apoB mRNA editing in the fasted/refed rat model of hepatic editing (where fasting inhibited editing and refeeding with a low fat, high carbohydrate diet induced editing ) and in the hyperinsulinemic state [26; 27]. Alternatively alterations in APOBEC-1 expression did not occur when hepatic editing activity was enhanced in hypothyroid  or ethanol-fed rats  or in ethanol treated rat primary hepatocytes cultures or McArdle cells [29; 30]. Instead, ACF serine phosphorylation was stimulated by insulin or ethanol which resulted in an increased nuclear retention of ACF [10; 29; 31], an enhanced interaction of ACF with APOBEC-1 promoting editosome assembly and increased editing activity . Phosphorylation of ACF did not affect its interaction with apoB mRNA. The data presented in this report suggested that similar regulatory mechanisms controlling the functional interaction of APOBEC-1 with ACF might be operational during enterocyte differentiation. Subsequent to differentiation, intestinal apoB mRNA editing has >85% efficiency suggesting a constitutive ability of enterocytes to maintain activated editing factors. The mechanisms have importance for all tissues that edit and will be the subject of our ongoing investigations.
The authors thank Dr. Mark Sowden and David Lehmann for technical assistance and thoughtful discussion and Jenny Smith for the preparation of figures. This work was supported by Public Health Services Grant DK43739 and Air Force Grant F49620.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.