Previously, we associated two mutations with the diseases X-linked congenital spinal muscular atrophy (SMAX2) and a syndromic form of X-linked mental retardation (MRXS10). These mutations are synonymous, meaning that they do not influence the amino acid sequence of the respective protein. They are localized in exon 15 of the gene ubiquitin activating enzyme 1 (UBA1) and exon 5 of the hydroxyacyl-coenzyme A dehydrogenase-II gene (HADH2). As quantitative PCR (qPCR) revealed, these mutations caused a strong decrease of the mRNA expression of the both genes. Herewith, we show that each mutation affects the functionality of exonic splicing enhancers (ESE) critical for proper splice site usage. Firstly, the wild type exonicUBA1 and HADH2 sequences appeared to promote pre-mRNA splicing within an ESE dependent reporter, whereas both mutant sequences exhibited an impaired recognition of the respective 5´ splice site (5´ss). Secondly, RNA pull-down assays revealed reduced binding of the splicing regulatory SR-proteins SRSF2 and SRSF6 to the mutant RNA sequences of UBA1 and HADH2. Remarkably, we found that the mutant HADH2 pre-mRNA interacts with hnRNP F/H known to repress splicing from exonic positions. Therefore, we propose that functional ESEs in UBA1 and HADH2 are disabled by synonymous mutations and, in case of HADH2, an hnRNP F/H dependent exonic splicing silencer (ESS) is created. Our findings emphasize the substantial influence of synonymous mutations on gene expression and stress their impact on comprehensive genetic diagnostics and counseling.

•    Gene expression
•    Splicing
•    Synonymous mutation
•    Neuronal diseases
•    Splicing regulatory element

Lenski C, Erkelenz S, Ramser J, Widera M, Peter JO, et al. (2015) Synonymous Mutations in the X-Linked Disease Genes UBA1 and HADH2 Affect Binding of the Splicing Regulatory Proteins SRSF2, SRSF6 and hnRNP F/H. JSM Genet Genomics 2(1): 1007.

MRXS10: X-Linked Mental Retardation; SMAX2: X-Linked Spinal Muscle Atrophy; SRE: Splicing Regulatory Element; ESE: Exonic Splice Enhancer; ESS: Exonic Splice Silencer; 5´Ss: 5´-Splice Site; qPCR: Quantitative PCR.

After a systematical screening for mutations associated with two severe X-linked neurological disorders, SMAX2 (MIM #301830) and MRXS10 (MIM #300220), by sequence analysis, we identified two distinct synonymous mutations in the genes UBA1 (NM_003334) and HADH2 (NM_004493) in the DNA taken from patients of six SMAX2 families and the MRXS10 family. In the SMAX2 patients, the mutation c.1731C>T in exon 15 of UBA1 [1] was detected, and c.574C>A in exon 5 of HADH2 in the MRXS10 family [2] (exons are numbered like in (1) and (2)). Both mutations are associated with considerably reduced mRNA expression levels of the affected genes within patient cells compared to healthy male controls as we showed by qPCR. In both cases the expression of the full-length transcripts was reduced [1,2]. Additionally, the expression of transcript variants lacking the exon 5 containing mutation showed increased levels in case of HADH2 [2].Considering also the in-silico analyses using the ESE finder tool, which predicted different ESE motifs here [1], it seemed to be obvious that important splicing regulatory elements (SRE) are targeted by these synonymous mutations. Now we present the identification of exonic SREs in both genes, UBA1 and HADH2, whose malfunctioning explains the observed reduced expression of the full-length transcripts. SREs modulate the recognition of splice sites in addition to their intrinsic strength [3,4]. The character of this modulation depends on the relative position of a SRE to the respective splice site [5]. Here, we could find two effects of the synonymous mutations in UBA1 and HADH2, the loss of function of ESEs in both genes as well as the creation of an ESS in HADH2.

The UBA1 and HADH2 wild type and mutant sequences (Metabion) as well as a neutral splicing regulatory sequence [6] and a known splicing enhancer as positive control [7] were inserted each into an enhancer-dependent HIV-1 based splicing reporter for 5’ss activation upstream of the 5’ss (EcoRI and SacI sites, Figure 1A) [8,9].

Figure 1: Analyses of the ESE/ESS activity of UBA1 and HADH2

 (A) The DNA oligonucleotides and the HIV-based splice enhancer dependent env-eGFP expression vector used in these studies.

(B-C) Analyses of the splice activity of the UBA1 and HADH2 constructs expressed in the env-eGFP-splice reporter by fluorescence microscopy and flow cytometry. As expression control pCL-dTom was co-transfected.

(B and C) Analyses of the ESE activity

HeLa cells were transiently transfected (TransIT-LT1, MIRUS) with the respective splicing reporters together with pCL-dTOM and pXGH5 containing the human growth hormone factor GH1 for qPCR analysis as transfection controls [5]. Twenty-four hours post transfection, HeLa cells were analyzed by direct fluorescence microscopy (Nikon ECLIPSE TS100-F, Nikon NIS-Elements imaging software suite) and flow cytometry using FACSCalibur (BD, Franklin Lakes).

For qPCR, total RNA was extracted and reverse transcribed into cDNA (DNAse I, Roche; Superscript III, Invitrogen). qPCR was performed on the LightCycler 1.5 (Roche) using LightCycler DNA Master SYBR Green I kit (Roche) and primers specific for eGFP and GH1 as control gene (primers available on request). Each analysis was repeated three times independently.

RNA pull-down experiments were performed in three technical replicates according to a well-established protocol [5,10] using agarose-beads coupled to RNA oligonucleotides (Metabion; adipic acid dihydrazide-Agarose, Sigma) (Figure 2A),

Figure 2:

Analyses of the interactions between RNA fragments of UBA1 and HADH2 and SR proteins

 (A) Analyses of binding proteins: The RNA oligonucleotides used in the RNA pull down experiments to interact with proteins from a HeLa nuclear extract. The synonymous mutations in the fragments of UBA1 and HADH2 are written in Italics.

(B-D) Western blots of proteins interacting with the RNA fragments of UBA1 and HADH2 after the RNA pulldown reaction.

representing either wild type or mutant sequence fragments of UBA1 and HADH2. These complexes were incubated in a HeLa-nuclear extract (Cilbiotech) for protein interaction. After separation on a polyacrylamide gel (12%), mass spectrometry was performed to identify the bound proteins (ZMMK, Cologne, Germany).

For Western blotting analyses, anti-SRSF2 (ab28428, Abcam), anti-SRSF6 (WH0006431M1, Sigma-Aldrich), anti-hnRNP F/H (ab10689, Abcam), anti-SRSF1 (sc-10255, Santa Cruz), anti SRSF7 (sc-10245, Santa Cruz) and anti-hnRNP-A1 (sc-56700, Santa Cruz) were used.

The two described variants c.574C>A (HADH2) and c.1731C>T (UBA1) were submitted to the LOVD-database (www.LOVD.nl).

Using an enhancer-dependent, HIV-1 based splicing reporter, the UBA1 and HADH2 derived sequences (Figure 1A) function as strong ESEs critical for the required 5´ss recognition by the spliceosome [9, 11]. As observed by direct fluorescence microscopy, both wild type exon-derived sequences could promote activation of the splicing reporter 5’ss resulting in a high expression of eGFP comparable to the positive control (Figure 1 B-C). In contrast, the mutant sequences of UBA1 and HADH2 showed a decreased 5’ss activation resulting in lower levels of eGFP signals (Figure 1 B-C). For both sequences, flow cytometry confirmed the reduced enhancer-dependent eGFP expression: Compared to wild type fragments and the positive control, the expression levels of eGFP were below 60% and 25%, respectively (Fig. 1 B-C, bar charts). These results were in agreement with qPCR analysis on mRNA level, where the eGFP expression was reduced to 40%, if the expression was regulated by the mutated UBA1 exon 15 sequence and to about 18% in case ofthe mutated HADH2 exon 5 sequence (data not shown). Therefore, both exonic wild type sequences of UBA1 and HADH2 contain ESE activities and each of these ESEs were impaired by the synonymous mutations c.1731C>T in UBA1 and c.574C>A in HADH2.

Next, performing RNA pull-down and mass spectrometric analyses, we could identify the known splicing factors SRSF1, SRSF2, SRSF6, SRSF7, hnRNP A1 and hnRNP F/H as candidate binding proteins of the UBA1 and HADH2 derived RNA substrates [12]. Western blotting analyses showed selective capabilities of the UBA1 and HADH2 fragments to bind either SR-proteins or hnRNPs. Using a specific anti-SRSF2 antibody, we demonstrated that binding of SRSF2 to the UBA1 exon 15 fragment was decreased in case of the c.1731C>T carrying fragment (Figure 2B). As detection by the anti-SRSF6 antibody revealed, the wild type HADH2 exon 5 fragment could be shown to interact with SRSF6, while this interaction was impaired by the c.574C>A mutation (Figure 2C). Moreover, the mutated HADH2 exon 5 fragment showed an increased binding of hnRNP F/H (Fig. 2 D), known to repress splice site recognition when positioned upstream of a 5´ss [5,10]. Western blotting analyses using antibodies against SRSF1, SRSF7 and hnRNP A1 did not show any differences between wild type and mutant fragments (data not shown).

The combination of different analysis methods, an enhancerdependent splicing reporter based analysis and RNA pull-down experiments, enabled a clear view of functional SREs in UBA1 and HADH2 and their interaction with the respective SR-protein. Two synonymous single base pair mutations within exons 15 and 5 of UBA1 and HADH2 could be shown to impair the ESE activities resulting in strongly reduced expression levels in an enhancerdependent HIV-1 based splicing reporter. In accordance with these reduced expression levels, we could demonstrate by RNA pull-down analyses that the two well-known ESE-binding SR-proteins, SRSF2 and SRSF6, could not bind anymore to the mutated exonic fragments of UBA1 and HADH2, efficiently. For the synonymous mutation inHADH2, we showed an even twofold effect on splice site recognition: The impairment of a so far unknown ESE and a de novo creation of an hnRNP F/H-dependent ESS which has numerously shown to interfere with functional spliceosome assembly, in this case across the HADH2 patient target intron. Herewith, the clearly specifiable characteristics of an inactivated enhancer and a newly formed silencer could be illustrated exemplary.

Our studies illustrate the severe outcome of single synonymous mutations affecting normal pre-mRNA processing. So far unknown exonic splicing enhancers within exons 15 and 5 of UBA1 and HADH2, respectively, could be found to be impaired by synonymous mutations detected in patients suffering from the severe neurological diseases SMAX2 and MRXS10. Apart from an enhancer inactivation, we could also show clear evidence for the de novo creation of an active hnRNP F/H-dependent silencer. These findings are exemplary as the underlying mechanisms are in accordance with shown mechanisms affecting the splicing regulation and the gene expression in case of other inherited diseases. Therefore, they emphasize the importance of analyzing the splicing modified at the RNA level due to synonymous mutations, especially in the context of comprehensive diagnostics and potential therapeutics.

1. Ramser J, Ahearn ME, Lenski C, Kemal O. Yariz, Heide Hellebrand, Michael von Rhein, et al. Rare Missense and Synonymous Variants in UBE, Playing a Key Role in the Ubiquitin-Proteasome System, are Associated with X-linked Infantile Spinal Muscular Atrophy. Am J Hum Genet. 2008; 82: 188-193.

2. Lenski C, Kooy RF, Reyniers E, Loessner D, Wanders RJ, Winnepenninckx B, et al. The reduced expression of the HADH2 protein causes X-linked mental retardation, choreoathetosis, and abnormal behavior. Am J Hum Genet. 2007; 80: 372-377.

3. Freund M, Hicks MJ, Konermann C, Otte M, Hertel KJ, Schaal H. Extended base pair complementarity between U1 snRNA and the 5’ splice site does not inhibit splicing in higher eukaryotes, but rather increases 5’ splice site recognition. Nucleic Acids Res. 2005; 33: 5112- 5119.

4. Nilsen TW, Graveley BR. Expansion of the eukaryotic proteome by alternative splicing. Nature. 2010; 463: 457-463.

5. Erkelenz S, Mueller WF, Evans MS, Busch A, Schöneweis K, Hertel KJ, et al. Position-dependent splicing activation and repression by SR and hnRNP proteins rely on common mechanisms. RNA. 2013; 19: 96-102.

6. Zhang XH, Arias MA, Ke S, Chasin LA. Splicing of designer exons reveals unexpected complexity in pre-mRNA splicing. RNA. 2009; 15: 367-376.

7. Cavaloc Y, Bourgeois CF, Kister L, Stévenin J. The splicing factors 9G8 and SRp20 transactivate splicing through different and specific enhancers. RNA. 1999; 5: 468-483.

8. Kammler S, Leurs C, Freund M, Krummheuer J, Seidel K, Tange TO, et al. The sequence complementarity between HIV-1 5’ splice site SD4 and U1 snRNA determines the steady-state level of an unstable env pre-mRNA. RNA. 2001; 7: 421-434.

9. Caputi M, Freund M, Kammler S, Asang C, Schaal H. A bidirectional SF2/ASF- and SRp40-dependent splicing enhancer regulates human immunodeficiency virus type 1 rev, env, vpu, and nef gene expression. J Virol. 2004; 78: 6517-6526.

10. Widera M, Erkelenz S, Hillebrand F, Krikoni A, Widera D, Kaisers W, et al. An intronic G run within HIV-1 intron 2 is critical for splicing regulation of vif mRNA. J Virol. 2013; 87: 2707-2720.

11. Cartegni L, Chew SL, Krainer AR. Listening to silence and understanding nonsense: exonic mutations that affect splicing. Nat Rev Genet. 2002; 3: 285-298.

12. Lin S, Fu XD. SR proteins and related factors in alternative splicing. Adv Exp Med Biol. 2007; 623: 107-122.