Alternative polyadenylation in a family of paralogous EPB41 genes generates protein 4.1 diversity

ABSTRACT

Alternative polyadenylation (APA) is a step in mRNA 3′-end processing that contributes to the complexity of the transcriptome by generating isoforms that differ in either their coding sequence or their 3′-untranslated regions (UTRs). The EPB41 genes, EPB41, EPB41L2, EPB41L3 and EPB41L1, encode an impressively complex array of structural adaptor proteins (designated 4.1R, 4.1G, 4.1B and 4.1N, respectively) by using alternative transcriptional promoters and tissue-specific alternative pre-mRNA splicing. The great variety of 4.1 proteins mainly results from 5′-end and internal processing of the EPB41 pre-mRNAs. Thus, 4.1 proteins can vary in their N-terminal extensions but all contain a highly homologous C-terminal domain (CTD). Here we study a new group of EPB41-related mRNAs that originate by APA and lack the exons encoding the CTD characteristic of prototypical 4.1 proteins, thereby encoding a new type of 4.1 protein. For the EPB41 gene, this type of processing was observed in all 11 human tissues analyzed. Comparative genomic analysis of EPB41 indicates that APA is conserved in various mammals. In addition, we show that APA also functions for the EPB41L2, EPB41L3 and EPB41L1 genes, but in a more restricted manner in the case of the latter 2 than it does for the EPB41 and EPB41L2 genes. Our study shows alternative polyadenylation to be an additional mechanism for the generation of 4.1 protein diversity in the already complex EPB41-related genes. Understanding the diversity of EPB41 RNA processing is essential for a full appreciation of the many 4.1 proteins expressed in normal and pathological tissues.

Introduction

EPB41-related genes are composed of a group of 4 paralogous genes^1-5 that together encode an impressive array of cytoskeletal proteins. The prototypical EPB41 gene is best known for encoding the protein known as protein 4.1R, which was originally identified as an 80-kDa component of the membrane skeleton of human red blood cells.⁶ Protein 4.1R is the founding member of the large family of proteins known as the band 4.1 superfamily, all of whose members contain a highly conserved region designated ‘the FERM domain’.⁷ The domain takes its name from the 4.1 (4 point one) and ERM (ezrin, radixin moesin) proteins in which it was discovered. FERM-containing proteins comprise a large number of proteins that associate with membrane proteins and with the cytoskeleton.⁸ Protein 4.1R and 3 homologues of 4.1R, namely 4.1N (abundant in neurons; encoded by the EPB41L1 gene), 4.1G (general distribution; encoded by the EPB41L2 gene) and 4.1B (abundant in brain; encoded by the EPB41L3 gene), constitute the protein 4.1 family.

Studies of EPB41 gene expression have shown that while the expression pattern of 4.1R in mature red cells is relatively simple, nevertheless, a rather complex array of 4.1R protein isoforms of varying sizes⁹ and different subcellular locations^8,10-18 exists in nucleated cells. The great diversity of protein 4.1R isoforms observed in nucleated cells mainly originates from the alternative splicing of the 4.1R pre-mRNA.^1,5,19,20 At least 10 internal coding exons can undergo alternative splicing, resulting in the generation of an extensive range of 4.1R isoforms that vary in their internal amino acid sequences. In addition, 2 groups of 4.1R isoforms varying in their N-terminal extensions can be generated if exon 2′ (17 nt) is retained or skipped. If exon 2′ is retained, the translation initiation site present in exon 2′ is used, while a second translation initiation site present in exon 4 is used if exon 2′ is skipped (Fig. 1A). As a consequence, 4.1R proteins containing or lacking 209 amino acids at their N-terminus are originated.^20-22 The prototypical erythroid protein 4.1R is translated from the second translation initiation site.⁵ More recent studies have revealed great complexity in the 5′ region of the EPB41 gene and shown evidence of coupling between transcription and alternative splicing that directly affects alternative splicing of exon 2′.²³ We recently reported an additional layer of complexity in the already complex EPB41 gene-regulation pathways based on our finding that alternative internal translation, more specifically an IRES-driven translation, is also involved in the generation of 4.1R isoforms lacking the 209 amino acid N-terminal extension.²⁴ All of these protein 4.1R isoforms contain a unique sequence domain, known as the C-terminal domain (CTD) that, in mammals, is encoded by exons 18–21.⁸ While the mechanisms responsible for the generation of 4.1R isoforms that vary in their N-terminal domains have been thoroughly analyzed, little is known about the contribution of RNA 3′-end processing in the EPB41- related gene family.

Figure 1.

Structure of the human brain EPB41 cDNAs cloned. (A) Schematic of the exon map of the protein 4.1R. Exons are coded as follows: blue, alternative; gray, constitutive; black, non-coding. The number of each individual exon is indicated at the top. Two translation-initiation sites at exon 2′ (ATG1) and exon 4 (ATG2) are indicated, as well as the stop codon (TGA) at exon 21. 1A, 1B and 1C represent alternative transcriptional promoters, the use of which results in retention or skipping of exon 2′ and, concomitantly, translation from the ATG1 or ATG2, respectively. Exons encoding the CTD region of the 4.1R protein are indicated. (B) Comparison of the sequence of the EPB41 gene at exons 17 to 18 with that of the BC039079 cDNA. The stop codon (TGA) in blue, the polyadenylation signal hexamer (AATAAA) in red, the cleavage site (TA) in green and the poly(A) tail in the BC039079 cDNA are indicated. (C) Exon composition of the 4 cDNAs isolated from human brain. The red box represents the new terminal exon, which we call 17′, that results from intronic APA. The amino acid sequence at the C-terminus of the protein encoded by exon 17′ is shown.

Maturation of mRNA involves multiple processing steps, including capping, splicing and polyadenylation. mRNA polyadenylation is a 2-step reaction,^25,26 involving specific endonucleolytic cleavage at the polyadenylation site (poly(A) site) and subsequent polymerization of an adenosine tail in which several proteins participate in a cooperative manner.²⁷ Mammalian pre-mRNAs contain 3 primary sequence elements that define the poly(A) site and 2 auxiliary sequence elements that enhance and regulate the 3′-end processing reaction. The 3 primary sequence elements consist of the polyadenylation signal (PAS) constituted by the hexamer AAUAAA or less frequently AUUAAA and other hexamer variants,^28,29 the cleavage site, and the G/U-rich downstream element.³⁰ The 2 auxiliary sequence elements consist of an upstream element, often consisting of a U-rich element, and a downstream element that is generally a G-rich sequence.³⁰ In vertebrate pre-mRNAs, 70% of the cleavage sites are located on the 3′ side of an adenosine residue, with a nucleotide preference of A > U > C > > G. The nucleotide preceding the cleavage site is cytosine in more than 50% of the pre-mRNA sequences examined, making CA the optimal cleavage site.^31,32

Over half of all human pre-mRNAs have multiple poly(A) sites,³³ implying that alternative polyadenylation (APA) is a widespread phenomenon that generates transcript isoforms with alternative 3′ ends. These can code for the same protein, the most frequent types of APA, or for C-terminally modified proteins, the least frequent types.^34,35 Indeed, APA can significantly contribute to the complexity of the proteome in the cell.

In this study we have isolated 4 EPB41 cDNA species from human brain that diverge in their 3′-end sequence from prototypical EPB41 cDNAs because they lack exons 18–21, thereby encoding unique C-terminal-truncated 4.1R isoforms that lack the prototypical CTD characteristic of proteins 4.1R. We have examined whether 3′-end-truncated EPB41 transcripts exist in human tissues other than brain, analyzed their expression relative to prototypical EPB41 mRNAs, explored the evolutionary conservation of 3′-end RNA processing, more specifically, alternative polyadenylation, in the EPB41 gene among various mammals and examined whether the EPB41L1, EPB41L2 and EPB41L3 genes conserve this APA in human tissues. Our findings show APA to be an additional mechanism that generates protein diversity in the already complex EPB41-related genes.

Results

APA generates EPB41 transcripts encoding C-terminal-truncated 4.1R proteins in human brain

By analyzing the genetic database of the Mammalian Gene Collection (MGC) from the National Institutes of Health (NIH), we detected a full-length EPB41 cDNA isolated from human brain (GenBank accession number: BC039079), which differed in its 3′ region from all described previously EPB41 cDNAs. While the stop codon of prototypical EPB41 cDNAs is situated within exon 21 (exons are numbered according to Huang et al.³⁶), the BC039079 cDNA lacked exons 18 to 21. Inspection of the 3′-end of the BC039079 cDNA revealed that it was made up of 62 bp of what, until now, has been considered a part of intron 17, followed by a poly(A) tail (Fig. 1). Analysis of the extra 62 bp allowed the identification of special features within this sequence. At nucleotides 42 to 47 there is an AATAAA hexamer sequence that is characteristic of prototypical PAS. Downstream of the PAS hexamer, at nucleotides 61 and 62, there is a TA pair followed by a poly(A) tail comprising 85 As. Thus, this TA pair acts as a cleavage site, although it is less optimal than the CA pair identified as a prototypical cleavage site.³¹ The presence of a stop codon, TGA, at nucleotides 19 to 21 indicates that this EPB41 transcript encodes a unique C-terminal-truncated 4.1R isoform.

To analyze further the complexity of 3′-end-truncated EPB41 transcripts in human brain, we performed RT-PCR analysis from commercial human brain RNA using specific sense and antisense oligonucleotides for exon 2′ and intron 17, respectively (see Methods). EPB41 cDNAs of the predicted 2-kb size were amplified, cloned and sequenced. The resulting 11 cDNA clones were classified into 4 groups according to their exonic composition (Fig. 1C and Table 1). The variation between the 4 groups consisted of the presence or absence of exons 14, 15 and 16 due to alternative splicing events. Five clones contained exon 15, reported to be specific to brain tissue,³⁶ and 5 contained exon 16, which is only selectively included in a few cell types and tissues.^20,21,36,37 The 4 clones constituting group 2 were identical to BC039079. It is of note that some of the clones from groups 3 and 4, had a deletion of 3 nucleotides, CAG (nt 1637–1640 of the NM_001166005.1 coding sequence) at the 5′-end of exon 13 and concomitantly the exon-13-encoded sequence in these 2 4.1R isoforms gives rise to sequences of 74 instead of 75 amino acids. A similar deletion was also observed for prototypical EPB41 cDNAs encoding the CTD that resulted from the use of a different cryptic acceptor site for alternative splicing in exon 13, thereby eliminating a triplet.³⁸ The length of the amino acid sequence encoded by our panel of 3′-end-truncated EPB41 cDNAs and their predicted molecular weights is shown in Table 1.

Table 1.

Characterization of EPB41 cDNAs generated by APA in human brain.

Group*	Total number of clones	Exon 14	Exon 15	Exon 16	Clones with CAG deletion	Predicted MW (Da)	Number of amino acids
1	1	+	+	+	0	82,922	734
2	4	+	−	+	0	81,235	720
3	4	+	+	–	1	80,325	713
						80,250	712
4	2	−	−	−	1	76,580	680
						76,485	679

^*The 4 groups of full-length isolated cDNAs are classified on the basis of their exonic composition; (+) presence, and (−) absence of alternative exons 14, 15 and 16. Clones with a CAG deletion at the 5′-end of exon 13 (E13) are indicated.

The results presented so far indicate that there exists in human brain a set of 3′-end-truncated EPB41 transcripts that contain an exon comprising 18 coding nucleotides and that is not present in prototypical EPB41 transcripts; we will refer to this as exon 17´. These transcripts encode a set of unique C-terminal-truncated 4.1R isoforms in which the characteristic CTD (22–24 kDa) of prototypical proteins 4.1R is replaced by a specific 6-amino acid sequence, VSTLST, at their C-terminus (Fig. 1C).

All prototypical 4.1R proteins contain the CTD but can vary in their N-terminal extensions. They are generated by 2 groups of transcripts: those containing the translation initiation site, AUG1, present in exon 2′, and those that, as a result of an alternative splicing event that eliminates exon 2′, lack the AUG1 and hence use a second AUG2 present in exon 4 as their translation initiation site.²¹ To explore whether 3′-end-truncated EPB41 transcripts lacking the AUG1 but containing the AUG2 exist in human brain, we used the same conditions and reverse primers used to clone the 3′-end-truncated EPB41 cDNAs containing the ATG1 described above (see Methods). As a sense oligonucleotide we used the same one that worked to successfully amplify prototypical EPB41 cDNAs lacking the ATG1 but containing the ATG2 translation initiation site from Molt-4 T cells.³⁹ We did not obtain any cDNA products. Together, these results indicate that 3′-end-truncated transcripts containing the AUG1 do exist in human brain, but the type of transcript that lacks the AUG1 and uses the AUG2 as its translation initiation codon does not appear to be present.

APA generates 3′-end-truncated EPB41 transcripts in various human tissues

To investigate whether human tissues other than brain express 3′-end-truncated EPB41 transcripts, we performed RT-PCR using commercial human mRNAs and specific oligonucleotides covering the sequence from exon 11 to exon 17′. The results obtained are summarized in Table 2. All of the 10 human tissues analyzed (bone marrow, kidney, liver, prostate, salivary gland, skeletal muscle, small intestine, spleen, testis and thymus) expressed 3′-end-truncated EPB41 transcripts, indicating that this type of mRNA is not specific to brain. It is of note that exon 15 was only expressed in human brain (compare Tables 1 and 2). Three of the 10 tissues analyzed (liver, kidney and small intestine) expressed more than one 3′-end-truncated EPB41 transcript species, varying with respect to the expression of exons 14, 15 and 16. The other 7 tissues expressed only one EPB41 transcript species lacking exons 14, 15 and 16 (5 of the 7 tissues) or lacking exons 15 and 16 (2 of the 7 tissues). Similar to the results obtained in brain, the CAG triplet at the 5′-end of exon 13 was eliminated in some of the transcripts; this event occurred more generally in liver (Table 2). The poly(A) site studied here has been documented in the APA database designated as polyA_DB 2⁴⁰ with the entry number Hs 175437.1.16. It corresponds to position 29212349 in chromosome 1 in the assembly NCBI35/hg17.

Table 2.

Characterization of EPB41 cDNAs generated by APA in various human tissues.

Tissue	Total number of clones	Exon 14	Exon 15	Exon 16	Clones with CAG deletion
Liver	2	+	−	−	1
	2	−	−	+	1
	2	−	−	−	2
Kidney	1	+	−	−	0
	2	−	−	−	0
Small intestine	1	−	−	+	0
	3	+	−	−	0
Skeletal muscle	4	−	−	−	0
Bone marrow	3	−	−	−	0
Prostate	3	+	−	−	0
Salivary gland	3	+	−	−	0
Spleen	3	−	−	−	1
Testis	2	−	−	−	0
Thymus	3	−	−	−	1

To compare the expression of 3′-end-truncated EPB41 mRNAs relative to prototypical EPB41 mRNAs we performed RT-qPCR assays using RNA from various human tissues. The results in Fig. 2 show that this ratio varies between 24 and 38%, depending on the source of the RNA.

Figure 2.

Relative expression of EPB41 mRNA species. RNA from the indicated human tissues were subjected to RT-qPCR for 3′-end-truncated and prototypical EPB41 mRNAs and their relative expression determined. Black, prototypical EPB41 mRNA; white, 3′-end-truncated EPB41 mRNA. Bars indicate the standard deviation of triplicate samples.

Comparative analysis of the EPB41 exon 17′ in different mammal species

Inspection of the sequences of the EPB41 gene in different species suggested that 3′-end-truncated EPB41 transcripts might exist in species other than humans. Fig. 3A shows the nucleotide coding sequence of exon 17′ and the amino acid sequence encoded by this exon in human, gorilla, dog, cow, rat and mouse. The latter sequences, which vary between 6 and 8 amino acids in length, are highly homologous with that of human. Fig. 3B shows the complete sequence of exon 17′ for human, gorilla and cow and indicates that the homology is very high, not only at the nucleotide level but also with respect to PAS organization. In these species, the PAS hexamer is 10–20 nt downstream of the stop codon and a TA signal is 13–16 nt downstream of the canonical PAS hexamer. All these results suggest that APA occurs in species other than humans.

Figure 3.

Sequence comparison of nucleotides in exon 170 of the EPB41 gene in various mammals. (A) Comparison of the nucleotide coding sequence of exon 17′ (left) and of the amino acid sequence encoded by it (right) in various mammals. The stop codon, TGA, is shown in blue. The amino acid sequences were aligned using Clustal Omega and Jalview with the Blosum62 coloring scheme, whereby gaps are colored white, matching residues are in dark blue and non-matching but positively scored residues are in light blue.⁵⁵ (B) Coding and non-coding nucleotide sequences of exon 17′ of human, gorilla and cow are shown. The PAS hexamer, AATAAA, is shown in red, and the sites of mRNA cleavage and addition of the poly(A) tail are indicated in green. Asterisks indicate positions with a single, fully conserved residue; colons indicate conservation between groups of strongly similar properties – a score of >0.5 in the Gonnet PAM 250 matrix; periods indicate conservation between groups of weakly similar properties – a score of ≤ 0.5 in the Gonnet PAM 250 matrix.

Comparison of the proximal and distal poly(A) sites in the human EPB41 gene

Two functional PAS have been identified in EPB41 mRNAs: the distal one, in the 3′-most exon 22, used for generating the prototypical proteins 4.1R containing the CTD,³⁶ and a proximal one, located in exon 17′, analyzed in this study, used for generating 4.1R isoforms lacking the CTD. Fig. 4 shows the sequences of the proximal and distal PAS characterized for the EPB41 gene. Interestingly, both EPB41 PAS correspond to the strong canonical hexamer, AATAAA. To identify EPB41 cDNAs/mRNAs with poly(A) tails and learn about EPB41 mRNA cleavage sites associated with these 2 AATAAA hexamers, we examined the Evidence Viewer database. The mRNA cleavage site in exon 17′ is at the 3′ site of a TA pair which is 15 nt downstream of the PAS hexamer, whereas in the 3′-most exon 22 there are 3 tandem cleavage sites 8, 14 and 20 nt downstream of the AATAAA; the most 5′ one corresponds to the canonical CA pair and the most 3′ one corresponds to the weak TC site (Fig. 4). Interestingly, the latter cleavage site has been identified in the EPB41 cDNAs encoding the best-characterized proteins 4.1R.

Figure 4.

Polyadenylation sites in human EPB41 cDNAs. Nucleotide sequences 3′ to the PAS hexamer, AATAAA (underlined), in exon 17′ (proximal site) and exon 22 (distal site) of human EPB41 cDNAs. The corresponding regions of human EPB41 genomic DNA are shown. The sites of mRNA cleavage and addition of the poly(A) tail are indicated by arrowheads. In addition to the cDNAs indicated in the figure, AL8334831 and AK1727961.1 (a) AB209649.1 (b) NM.203343.2, NM.203342.2, NM.001166007.1 and NM.004437.3 (c) cDNAs with the same 3′ end were also identified in the Evidence Viewer database.

APA generates 3′-end-truncated EPB41L2, EPB41L3 and EPB41L1 mRNAs only in specific human tissues

We analyzed genetic databases to examine conservation of the polyadenylation sequences in the EPB41L2, EPB41L3 and EPB41L1 genes and noticed that the 3′-end RNA processing observed for the EPB41 gene could occur in the EPB41L2, EPB41L3 and EPB41L1 genes. To confirm this, we amplified cDNA fragments spanning from either exon 11 (for EPB41L1 and EPB41L3) or exon 12 (for EPB41L2) to exon 17′ (exons were numbered according to the homologous exons in the EPB41 gene). For the EPB41L2 gene (protein 4.1G), in 10 of 11 tissues analyzed, we detected one 3′-end-truncated mRNA species, all of which contained E14, E15 and E16 (Table 3), hence indicating that 3′-end RNA processing is not exclusive to the EPB41 gene. We found no 3′-end-truncated transcript for the EPB41L2 gene in bone marrow. A major difference between the results obtained for the EPB41 and EPB41L2 genes was noticed. While more than one species of 3′-end-truncated mRNAs was expressed in some of the tissues analyzed for the EPB41 gene (Table 2), we detected only one mRNA species, of the same exonic composition, in all of the 10 tissues positive for the EPB41L2 gene (Table 3). Our data also show that 3′-end-truncated mRNAs are produced for the EPB41L3 (protein 4.1B) and EPB41L1 (protein 4.1N) genes, although in a very restricted manner. Indeed, we only observed 3′-end-truncated mRNAs in brain and bone marrow for the EPB41L3 gene and in brain for EPB41L1. Only one mRNA species was identified for these 2 genes, with the exception of 2 EPB41L3 mRNA species detected in brain (Table 3).

Table 3.

Characterization of EPB41-related cDNAs generated by APA in various human tissues.

	Positive tissues*	Total number of clones	Exon 14	Exon 15	Exon 16
EPB41L2/ 4.1G	All except bone marrow	28	+	+	+
EPB41L3/ 4.1B	Brain	1	−	+	+
		2	−	+	−
	Bone marrow	2	−	+	−
EPB41L1/ 4.1N	Brain	3	+	+	+

^*The tissues analyzed for expression of the 3 EPB41-related genes (EPB41L2, EPB41L3 and EPB41L1) were: brain, liver, kidney, small intestine, skeletal muscle, bone marrow, prostate, salivary gland, spleen, testis and thymus. Only the tissues in which APA was detected are indicated.

Discussion

APA is recognized as an important level of gene regulation. The simplest and most frequent form of APA is tandem 3′UTR APA, which involves the occurrence of alternative poly(A) sites within the same terminal exon and thereby generates multiple isoforms that differ in the length of their 3′UTR without affecting the protein encoded by the gene. Three other types of APA event potentially affect the coding sequences in addition to the 3′UTRs: alternative terminal exon APA, in which alternative splicing generates isoforms that differ in their last exon; intronic APA, which contains the poly(A) site in an intron and extends an internal exon, converting it to a 3′ terminal exon known as a composite terminal exon; and internal exon APA, which involves premature polyadenylation within the coding region. These 3 types of APA contribute to the complexity of the transcriptome by generating isoforms with differing coding sequences.^{35, 41}

In this study we show that the use of a poly(A) site located in intron 17 in the EPB41 gene results in short human mRNAs encoding protein 4.1R isoforms that lack the CTD characteristic of prototypical 4.1R proteins. The intronic APA experienced by the EPB41 pre-mRNA involves cleaving at the intronic poly(A) site, extending the internal composite exon 17/17′ and making it the terminal one. An additional consequence of the extension of composite exon 17/17′ is that this type of 4.1R isoforms contains a distinct short amino acid sequence at its C-terminus, unlike the prototypical 4.1R isoforms. APA is not specific to the EPB41 gene, as was observed in the other 3 paralogous EPB41 genes. However, this mechanism is more general for the EPB41 and EPB41L2 genes than for the EPB41L1 and EPB41L3 genes, since 3′-end-truncated mRNAs were detected in all of the 11 tissues analyzed (for EPB41), in 10 tissues (for EPB41L2) or in 2 (for EPB41L3) and one (for EPB41L1) of the tissues. It should be noticed that brain is the only tissue that experiences APA for the 4 paralogous EPB41-related genes.

In mRNAs containing 2 or more poly(A) sites, the proximal sites tend to use variant signals and are therefore generally weaker, whereas the most distal ones tend to use the canonical PAS and are usually strong.⁴¹ Interestingly, the proximal and distal sites in EPB41 both contain the canonical PAS hexamer, AATAAA, although the site of mRNA cleavage is less optimal for the proximal site, compared with the most 5′ site in the distal one. A large number of human poly(A) sites in the 3′-most exon are known to have more than one cleavage site and the distance between the 5′-most cleavage site and the others is less than 24 nt.³³ This is the case for the 3′-most exon 22 of EPB41 in which we found heterogeneous poly(A) cleavage at 3 sites, occurring within 12 nt after the 5′-most cleavage site. By contrast, in intronic poly(A) sites, the most frequent finding in human genes is one cleavage site.³³

Erythroid mRNAs different from those of the EPB41 genes are also subject to 3′-end processing. For instance, APA at the 3′-untranslated region of ANK-1 accounts for 2 different-sized erythrocyte ANK-1 mRNAs.⁴² However, these 2 transcripts do not give rise to different proteins. By contrast, tissue-specific differential processing of 3′ ß-spectrin pre-mRNA results in structurally different ß-spectrin isoforms in red cells and other tissues.⁴³ The proximal poly(A) site seems to be used exclusively in erythroid cells and a default pattern of pre-mRNA processing relies on the use of the distal site. As a result, skeletal muscle ß-spectrin contains a different, longer carboxyl terminus than does erythroid ß-spectrin.

In the past, we learnt that 2 groups of 4.1R proteins varying in their N-terminus can be generated, depending on whether exon 2′ is spliced in or out.²¹ Our study shows that 2 groups of 4.1R proteins varying in their C-terminus can also be generated, depending on whether exon 17′ is retained. It is of note that although exons 2′ and 17′ are very small, their presence or absence nevertheless results in protein 4.1R isoforms containing or lacking large extensions at their N- or C-terminus, of 209 and 136 amino acids, respectively (Fig. 5). We did not detect transcripts that simultaneously retain exon 17′ and splice out exon 2′. This suggests that 4.1R proteins lacking both the N-terminus and the CTD appear not to exist. This study highlights that we must be cautious in interpreting experimental results when using antibodies generated against regions common to prototypical 4.1R proteins and CTD-lacking 4.1R isoforms, since we might be detecting the 2 types of isoforms in tissues in which both are expressed. We should also bear in mind that previous silencing experiments intending to eliminate the expression of all known 4.1R proteins mainly used siRNAs directed to regions downstream of the stop codon present in exon 21, which would not have been effective in silencing CTD-lacking 4.1R isoforms. As we were not previously aware of the existence of these isoforms, it is plausible that some of the effects observed might not have been as drastic as they should have been if all types of 4.1R isoforms (those with and without CTD) were silenced.

Figure 5.

Two groups of EPB41 mRNAs and encoded proteins generated by APA. (A) The scheme shows 2 mechanisms that generate variability at the 5′- and 3′-ends of EPB41 mRNAs, leading to N-terminal and C-terminal variations in proteins 4.1R. (B) Two groups of 4.1R proteins with different N-terminal or C-terminal extensions originated by alternative splicing ²¹ or by APA (this study), respectively. If exon 2′ is retained, the group NC (containing the N-terminal extension and the CTD) is translated from the ATG1; by contrast, if exon 2′ is skipped, the group ΔN-C (lacking the N-terminal extension and expressing the CTD) is translated from the ATG2. These 2 groups vary in their N-terminal regions. Intronic alternative polyadenylation results in the expression of exon 17′ and generation of 3′-end-truncated EPB41 mRNAs that originate the NΔC group (containing the N-terminal extension but lacking the CTD characteristic of prototypical proteins 4.1R). The ΔNΔC group is not represented here as it was not detected in our study. U1 and U2, represent unique variable regions in proteins 4.1R; FERM, represents the membrane-binding domain; SABD, represents the spectrin actin binding domain and CTD, represents the C-terminal domain. The regions encoded by exons 2′ and 17′ are indicated in red.

The CTD of 4.1R is involved in binding several proteins (the nuclear mitotic apparatus protein NuMA, the glutamate receptors GluR1 and GluR3, the peptidyl-prolyl isomerase FKBP13, among others)^44-47 that will not be bound to the set of CTD-lacking 4.1R proteins. It will be interesting to bring to light the function(s) that the latter isoforms play in the tissues in which they are expressed.

Defects in the EPB41-related genes have been shown to underlie human disease. For instance, hereditary elliptocytosis and hemolytic anaemia are associated with protein 4.1R deficiency in human red blood cells⁴⁸ and several studies suggest that EPB41 and its paralogues function as tumor suppressors.^49-51 Understanding the mechanisms by which expression of these genes is controlled and how the expression of their mRNAs is regulated will shed light on the generation of protein 4.1 diversity in the already complex EPB41-related genes.

Materials and methods

Full-length 4.1R cDNA cloning – Commercial human brain total RNA (Clontech) was used for cDNA synthesis. EPB41 cDNAs containing the ATG1 were amplified using a forward primer that hybridizes to exon 2′: 5′-CAACATCATGACAACAGAGA-3′ and a reverse primer that hybridizes to intron 17: 5′-GGTCATGTGGACAAAGTAC-3′. In parallel reactions, to amplify cDNAs lacking the ATG1 but containing the ATG2, we used a forward primer designated sinE2′: 5′-CCGCACCCAGCCCAGAGAAGAGTT-3′, previously used in our laboratory to amplify prototypical ATG1-lacking 4.1R cDNAs from Molt-4 T cells³⁹ and the reverse primer to intron 17 indicated above. Thirty-5 cycles of amplification were performed under the following conditions: denaturation for 30 s at 94°C; annealing for 30 s at 50°C; extension for 3 min at 68°C.

Reverse transcription for the 4 EPB41-related genes and PCR amplification of partial cDNA sequences – Commercial human total RNA Master Panel II (Clontech) containing total RNA of 11 tissues, including brain, were used for cDNA synthesis. Partial cDNA products for the 4 EPB41-related genes were amplified using forward primers directed to exon 11 (for EPB41L1 and EPB41L3) or exon 12 (for EPBL2) and reverse primers to intron 17 (exons were numbered according to the homologous exon in the EPB41 gene). The primers were as follows: forward 5′-GGAAAGTCTGTGTAGAAC-3′ and reverse 5′-GGTCATGTGGACAAAGTAC-3′ for EPB41; forward 5′-CTTCTCATTTACAAAGACAGAC-3′ and reverse 5′-GAAACTCATGCAATTTACCTAAG-3′ for EPB41L2; forward 5′-GGAAGGTCTGCATCGAGC-3′ and reverse 5′-CTAAGCAGCAATGGGGGAAG-3′ for EPB41L1 and forward 5′-GGAAAGTATGTGTTGAGC-3′ and reverse 5′-GCTCAGGTTAAACATAGG-3′ for EPB41L3. Thirty-5 cycles of amplification were performed under the following conditions: denaturation for 60 s at 94°C; annealing for 60 s at 55°C; extension for 120 s at 72°C.

cDNA cloning and sequencing – PCR amplification was resolved in a 1% agarose gel and single bands were purified with Wizard® SV Gel and PCR Clean-Up System (Promega) following the manufacturer's instructions. Nested PCR-purified products were cloned into the pCR4-TOPO vector using the TOPO-TA Cloning R kit (Life Technologies). cDNA sequences were confirmed by DNA sequence analysis (Macrogen).

RT-qPCR analysis for 3′-end-truncated EPB41 mRNAs and prototypical EPB41 mRNAs – Commercial human total RNA Master Panel II (Clontech) of 5 tissues were used for cDNA synthesis using oligo(dT)₂₀ primers and Superscript III retrotranscriptase (Invitrogen). To determine the levels of 3′-end-truncated EPB41 mRNA relative to prototypical EPB41 mRNA, absolute quantitative PCR was performed under the following cycling conditions: 5 s at 95°C + (5 s at 95°C + 5 s at 60°C) x 40 + (5 s at 60°C + 5 s at 95°C). Quantification was performed by real-time PCR using a CFX384 System (BioRad) in combination with SsoFast Eva Green (BioRad) and 0.5 μM of forward and reverse specific primers. For prototypical EPB41 cDNA, we used a forward (5′-GTGTGTATGTGTGTTTGTGTGAAGAAA-3′) and a reverse (5′-TGTAATTCTCTTCCCCCTCCTCA-3′) primer corresponding to exon 22. For 3′-end-truncated EPB41 cDNAs, we used a forward primer (5′-TAACATCAATGGGCAAATCCC-3′) corresponding to exon 17 and a reverse primer (5′-TCCCTCCATTTTCACAATTGGT-3′) complementary to exon 17′.

For both pairs of primers, specificity was tested by PCR of specific amplicons of each transcript. For prototypical EPB41 mRNA: 5′-AGTTGTGTGTATGTGTGTTTGTGTGAAGAAAAACAGACTCTGTCCAGGTAGAAATGGTGAGGAGGGGGAAGAGAATTACATTTCCAGGGTCAGAAACTTGGCAACAGT-3′ and for 3′-end-truncated EBP41 mRNA: 5′-GAACTCTTAACATCAATGGGCAAATCCCCACAGGAGAAGGAGTGAGTACTTTGTCCACATGACCAATTGTGAAAATGGAGGGAATAAATGTTTTTATGTATT-3′.

Amplicons were serially diluted to generate a calibration curve that was used to determine the number of copies of each mRNA. Data were analyzed with GenEx 5.3.7 software (Multid AnaLyses AB), normalized relative to total RNA and expressed as the percentage of 3′-end-truncated EPB41 mRNA relative to prototypical EPB41 mRNA.

Bioinformatic analysis – The human sequences of EPB41, EPB41L1, EPB41L2 and EPB41L3 genes were downloaded from the Ensembl database. The Evidence Viewer website from NCBI was consulted for cDNA identification. Translation of isolated EPB41 cDNA sequences was analyzed in silico using the Expasy Translate tool. Multiple cDNA sequence alignments of EPB41 homologues from different species were performed with the Clustal Omega program. Protein sequences were analyzed using the Jalview program.^52-54

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Acknowledgements

The authors wish to thank Dr Carlos Luque and Dr Javier Fernández-Chamorro for general discussions. We also appreciate the invaluable help of the Genomic and Bioinformatic Facilities of our Institute. We also thank Dr Phil Mason for revising the English language of the manuscript. This work was supported by grants BFU2011–22859 from the Ministerio de Educación y Ciencia, Spain, and S2010/BMD-2305-CM from the Comunidad de Madrid, Spain. L.R. was a predoctoral fellow of the Comunidad de Madrid.

Funding

This work was supported by grants BFU2011–22859 from the Ministerio de Educación y Ciencia, Spain, and S2010/BMD-2305-CM from the Comunidad de Madrid, Spain. L.R. was a predoctoral fellow of the Comunidad de Madrid.