A mutation-led search for novel functional domains in MeCP2

Jacky Guy; Beatrice Alexander-Howden; Laura FitzPatrick; Dina DeSousa; Martha V Koerner; Jim Selfridge; Adrian Bird

doi:10.1093/hmg/ddy159

A mutation-led search for novel functional domains in MeCP2

Guy, Jacky; Alexander-Howden, Beatrice; FitzPatrick, Laura; DeSousa, Dina; Koerner, Martha V; Selfridge, Jim; Bird, Adrian 2018-07-15 00:00:00 Most missense mutations causing Rett syndrome (RTT) affect domains of MeCP2 that have been shown to either bind methylated DNA or interact with a transcriptional co-repressor complex. Several mutations, however, including the C-termi- nal truncations that account for 10% of cases, fall outside these characterized domains. We studied the molecular conse- quences of four of these ‘non-canonical’ mutations in cultured neurons and mice to see if they reveal additional essential domains without affecting known properties of MeCP2. The results show that the mutations partially or strongly deplete the protein and also in some cases interfere with co-repressor recruitment. These mutations therefore impact the activity of known functional domains and do not invoke new molecular causes of RTT. The ﬁnding that a stable C-terminal truncation does not compromise MeCP2 function raises the possibility that small molecules which stabilize these mutant proteins may be of therapeutic value. Introduction polymorphisms are broadly distributed across the remainder of Mutations in the MECP2 gene cause the profound neurological the protein (Fig. 1). The implication that RTT mutational disorder Rett syndrome (1) (RTT, OMIM #312750). The gene prod- clusters mark critical domains is supported by biochemical uct, MeCP2 protein, is expressed in most cell types, but is espe- studies, which show that mutations within the Methyl-CpG cially abundant in neurons (2–4). Accordingly, loss of MeCP2 binding domain (MBD) disrupt DNA binding, whereas mutations function has the most profound effect in the nervous system, within the NCoR interaction domain (NID) prevent binding to with minimal phenotypic consequences in other tissues (3,5,6). the NCoR/SMRT co-repressor complexes (9–11). High-resolution Because of its association with intellectual disability in RTT and X-ray crystallographic structures of both these interactions other disorders, the MECP2 gene is frequently screened for support the view that MeCP2 bridges methylated sites on DNA mutations in clinical cases of developmental delay. This has de- with a histone deacetylase-containing co-repressor (12,13). ﬁned many amino acid changes as RTT-causing or as relatively Further evidence for the importance of these two domains benign or neutral variants that do not cause RTT (7,8). The comes from a recent study in which mice carried a pared down MeCP2 protein sequence is 95% identical between rodents and MeCP2 ‘minigene’, DNIC, consisting of only the MBD, a nuclear humans, suggesting stringent purifying selection throughout. localization signal and the NID (14). These mice survived for Nevertheless, RTT missense mutations are non-randomly dis- over a year with only mild symptoms. A simple interpretation is tributed in two prominent clusters, whereas apparently neutral that this interaction restrains transcription; a view supported Received: March 7, 2018. Revised: April 24, 2018. Accepted: April 26, 2018 V C The Author(s) 2018. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. 2531 Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 2532 | Human Molecular Genetics, 2018, Vol. 27, No. 14 ExAC variants 1 486 MeCP2 MBD MBD P225R P322L * * RTT missense CTDs RTT truncations Figure 1. Population variants and RTT mutations in the coding sequence of MeCP2. A schematic diagram of human MeCP2 protein shows the methyl-binding domain (MBD) and NCoR/SMRT interaction domain (NID). Single amino acid changes found in males in the ExAC exome sequence database are indicated above the protein and RTT-causing mutations are shown below. Only male ExAC changes are shown to exclude the possibility that pathogenic mutations in females could be masked by skewed X-inactivation. All RTT mutations have been shown to be absent in the parents in at least one case. Red box—range of deletion start points for C-terminal dele- tion (CTD) RTT frameshift mutations. The ‘non-canonical’ RTT mutations that are the subject of this study are marked with asterisks. by recent transcriptomic analyses of MeCP2-depleted mouse MBD and the NID. According to our rationale, new functional brain (15–17). domains would be expected to cause RTT without compromis- A signiﬁcant limitation of this model for MeCP2 function is ing either binding to DNA or NCoR recruitment. that several recurrent mutations, whose causal role in RTT is We initially selected the P225R and P322L mutations, which conﬁrmed by their absence in both parents, lie outside the MBD occur in a relatively small number of RTT cases (Table 1). Parental analysis has shown that these missense mutations and NID (Fig. 1) and are absent in DNIC mice (14). Most promi- nently, about 10% of all RTT cases involve a heterogeneous set arise de novo in the affected individuals (7), strongly suggesting that they are causative. We modeled these mutations by gene of small deletions in exon 4 of MECP2. The resulting frameshifts introduce premature translational termination often preceded targeting in mouse embryonic stem (ES) cells (Fig. 2a; Supplementary Material, Fig. S1a) to produce mouse alleles that by stretches of novel amino acid sequence of variable length differed from WT only at the patient mutation site and a single (18,19). As these C-terminal deletions (CTDs) are far down- loxP site remaining in intron 2 after deletion of the selection stream of the MBD and NID, they hint that there are other, as cassette (Fig. 2a; Supplementary Material, Fig. S1b–e). To deter- yet undetected, functional domains within MeCP2. The same ar- mine whether P225R and P322L cause RTT-like phenotypes in gument applies to two other recurrent RTT mutations, P225R mice, we injected mutant ES cells into blastocysts to create chi- and P322L, which lie outside the minimal regions required for P225R meras and bred them to establish Mecp2 (P225R) and DNA and NCoR binding. Here we use biochemical, cell biological P322L Mecp2 (P322L) mouse lines (Supplementary Material, Fig. and genetic methods to show that these mutations recapitulate S2a and b). P322L hemizygous males appeared normal at wean- RTT-like phenotypes in mice through either or both of two ing, but from 5 weeks of age they developed RTT-like symp- mechanisms: 1) by destabilizing MeCP2, so that only a fraction toms (20) that progressed rapidly (Fig. 2b), surviving to a median of the wild-type (WT) level remains; 2) by interfering with the age of 9 weeks (Fig. 2c). P322L males were also underweight ability of MeCP2 to recruit the NCoR complex to chromatin and compared to littermate controls (Supplementary Material, Fig. repress transcription. Importantly, we ﬁnd that one C-terminal S3a). These mice were subjected to a program of behavioral truncation is stable and functionally WT in a mouse sequence tests starting at 6 weeks old when the mutants were only mildly context, but unstable and RTT-causing when the protein is symptomatic. They showed signiﬁcantly decreased anxiety-like ‘humanized’. This demonstrates that the truncation itself behaviors in the elevated plus maze and open ﬁeld test only has a measurable functional consequence if it leads to de- (Supplementary Material, Fig. S3b and c) although their move- stabilization. From this and other data we conclude that these ment was not signiﬁcantly impaired (data not shown and ‘non-canonical’ RTT mutations do not identify previously unap- Supplementary Material, Fig. S3d). The wire suspension test and preciated functional domains. Moreover, our results highlight accelerating rotarod (Supplementary Material, Fig. S3e and f) ex- the crucial role played by the MeCP2 protein sequence in ensur- posed severe deﬁcits in motor co-ordination and learning. Thus ing stability. P322L males exhibit a severe phenotype mimicking that of Mecp2-null mice. Results As with P322L mice, P225R males developed Rett-like symp- toms at 6 weeks, but progression of symptoms was more grad- P225R and P322L mutations cause Rett-like phenotypes ual (Fig. 2d). Median survival was 50 weeks (Fig. 2e). P225R in mice males were underweight compared to WT littermates To test the hypothesis that the MBD and NID are key to MeCP2 (Supplementary Material, Fig. S4a). Despite their extended sur- function, we ﬁrst identiﬁed RTT mutations that apparently vival time (the longest of any Rett mutation modeled in mice so leave these domains intact. Next, we generated mouse models far), P225R mice showed a similar range of deﬁcits to P322L in to check that these recapitulated RTT-like phenotypes. Finally, behavioral tests. P225R males at 11 weeks of age appeared less we used cellular and molecular assays to ask whether these anxious than controls in the elevated plus maze and open ﬁeld mutations affect the functions that have been attributed to the test (Supplementary Material, Fig. S4b and c). They traveled a NID NID Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 Human Molecular Genetics, 2018, Vol. 27, No. 14 | 2533 (a) 12 3 4 Mecp2 genomic locus 30kb plasmid subclone targeting Neo Stop vector loxP loxP RTT mutation targeted genomic locus loxP (b) (c) P322L P322L WT WT null null 0 0 4 8 12 16 20 24 28 0 4 8 1216202428 Age (weeks) Age (weeks) (d) (e) 10 100 P225R WT 8 flox 80 null 6 60 P225R WT 4 40 null 2 20 0 0 4 8 12 16 20 24 28 32 36 40 44 48 52 0 4 8 1216202428323640444852 Age (weeks) Age (weeks) Figure 2. P322L and P225R mutant mice show RTT-like phenotypes and a reduced lifespan. (a) The mouse Mecp2 genomic locus, the targeting vector used to make P225R and P322L knock-in alleles in mouse ES cells, and the mutated genomic locus after Cre-mediated removal of the selection cassette. The region where the RTT mutations were introduced is marked with an asterisk (*). (b) P322L hemizygous male mice show a rapid onset of RTT-like phenotypes. Mean aggregate phenotypic score6SD plotted for each genotype, P322L n¼12 at 4 weeks, WT littermates n¼15. Representative Mecp2-null data are shown for comparison (n¼12 at 4 weeks). (c) Kaplan–Meier survival plot for P322L hemizygous males and WT littermates shown in (b). (d) P225R hemizygous male mice also display RTT-like phenotypes. Mean ag- gregate phenotypic score6SD plotted for each genotype, P225R n¼19 at 4 weeks, WT littermates n¼18. Representative Mecp2-null and Mecp2-ﬂox data are shown for comparison (ﬂox n¼9). (e) Kaplan–Meier survival plot for P225R hemizygous males and WT littermates shown in (d). Table 1. MECP2 mutations in this study. Amino acid and nucleotide locations are numbered according to the human e2 isoform. RettBASE fre- quency is expressed as a percentage of all Rett syndrome cases with an MECP2 mutation Mutation name Amino acid change Nucleotide change Mutation type RettBASE frequency MeCP2 domain R111G R111G c.331 A>G RTT missense 0.03 MBD P225A P225A c.673 C>G ExAC (male) n/a ID P225R P225R c.674 C>G RTT missense 0.55 ID R306C R306C c.916 C>T RTT missense 7.3 NID P322L P322L c.965 C>T RTT missense 0.26 CTD CTD1 L386HfsX5 c.1157–1197 D41 RTT del/frameshift 1.32 CTD CTD2 P389X c.1164–1207 D44 RTT del/frameshift 1.29 CTD MBD, methyl-CpG binding domain, ID, intervening domain (between the MBD and NID), NID, NCoR interacting domain and CTD, C-terminal domain. Aggregate Score Aggregate Score Percent survival Percent survival Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 2534 | Human Molecular Genetics, 2018, Vol. 27, No. 14 shorter distance in the open ﬁeld (Supplementary Material, Fig. heterochromatic foci containing high concentrations of methyl- S4d) and also demonstrated reduced motor co-ordination in the ated DNA (28). A mutation that disrupts the MBD, R111G, abol- wire suspension test (Supplementary Material, Fig. S4e), though ished binding of MeCP2 to methylated DNA and co-localization was lost. Both P225R and P322L (in addition to controls P225A mutants did not perform less well than controls on the acceler- ating rotarod (Supplementary Material, Fig. S4f). We conclude and R306C) retained the ability to bind to heterochromatic foci, indicating that MBD function is not affected in these mutants that both P322L and P225R male mice show phenotypes (Fig. 4a). that overlap with those of other mouse models carrying Mecp2 MeCP2 interacts with the co-repressor complex NCoR/SMRT mutations found in RTT (21–24). via subunits TBL1X and TBL1XR1. This interaction is disrupted by mutations in the NCoR/SMRT interaction domain, including P225R and P322L mutations differentially affect MeCP2 the common Rett mutation R306C. EGFP-tagged MeCP2 was levels expressed in HEK 293 cells and immunoprecipitated from 150 mM salt extracts using GFP-Trap beads (Supplementary To investigate the molecular basis of these RTT-like pheno- Material, Fig. S6a). NCoR/SMRT components NCoR1, TBL1X and types, we ﬁrst measured MeCP2 abundance. Strikingly, the HDAC3 co-immunoprecipitated with WT MeCP2 but not with amount of P322L protein in whole mouse brain was severely re- the R306C mutant or EGFP alone. Both P225R and P322L mutants duced (3% of WT littermates; Fig. 3a and b). Thus, despite WT P322L were able to immunoprecipitate the same NCoR/SMRT compo- amounts of Mecp2 mRNA (Supplementary Material, Fig. S5a nents as the WT protein (Fig. 4b), showing that these mutations and b), the level of P322L protein in brain is comparable to that Stop/y do not impair the interaction of MeCP2 with the co-repressor in Mecp2 mice, which also show a severe phenotype (22). In complex. The co-repressor mSin3a, which interacts relatively P225R mouse brain, MeCP2 protein was at 22% of the WT level weakly with MeCP2 (29), was immunoprecipitated by all mutant (Fig. 3c and d), which is signiﬁcantly more than in P322L MeCP2 proteins, including R306C (Supplementary Material, Fig. P225R brain. The amount of Mecp2 mRNA was not affected S6b). (Supplementary Material, Fig. S5b). Although P225R and P322L can still interact with NCoR and Reduced MeCP2 protein abundance was also observed in cul- methylated DNA separately, we wished to know if they could tured neurons derived from the mutant mouse ES cells (Fig. 3e– recruit the NCoR subunit TBL1X to heterochromatin, which g). Both P322L and P225R ES cells differentiated equally well into requires simultaneous binding to both macromolecules. To test neurons, as shown by NeuN expression (Fig. 3f) and displayed this, constructs expressing both EGFP-tagged MeCP2 and equal amounts of Mecp2 mRNA (Supplementary Material, Fig. mCherry-tagged TBL1X were transfected into mouse NIH-3T3 S5c). However, protein levels in both mutants were signiﬁcantly ﬁbroblasts (Supplementary Material, Fig. S7a–e). It has been lower than WT and WT þ lox controls (Fig. 3f and g). shown previously (30) that TBL1X-mCherry, which lacks a nu- Speciﬁcally, P225R showed a moderate reduction to around 60% clear localization signal, is normally excluded from the nucleus whereas P322L retained only 7% of the WT MeCP2 level. (Supplementary Material, Fig. S7b), but in the presence of WT EGFP-MeCP2 it is recruited to the methylation-rich 4’, 6-diami- dino-2-phenylindole (DAPI) spots in 80% of doubly transfected P225R and P322L impair NCoR recruitment to chromatin cells (Fig. 4c and d). As expected, the NID mutant R306C itself The drastic reduction in P322L MeCP2 explains the severe phe- localizes to the DAPI spots, but is unable to recruit TBL1X notype, but the P225R mutant is only partially depleted. (Fig. 4c). When P225R and P322L RTT mutants were tested in Notably, the level of P225R MeCP2 in whole brain is similar to this assay both showed impaired recruitment of TBL1X- that of the hypomorphic ﬂoxed Mecp2 allele (6,25–27)(Fig. 3c), mCherry to DAPI spots in comparison with WT MeCP2. On the which has a normal lifespan and much milder phenotypic pro- other hand, the population variant P225A performed as well as gression than P225R (25–27) [data from (25) shown for compari- WT MeCP2 in this assay (Fig. 4d; Supplementary Material, Fig. son in Fig. 2d, yellow symbols]. The difference between S7c). Thus the ability of both P225R and P322L to recruit NCoR/ phenotypic outcomes may indicate that these two mouse lines SMRT to methylated sites in the genome is impaired. Though simply lie on either side of a threshold for functionally sufﬁcient performed in non-neuronal cells, this test appears to measure a MeCP2 protein levels. Alternatively, the P225R mutation may af- property of MeCP2 that is functionally relevant, as Rett muta- fect MeCP2 function as well as abundance. To test the possibil- tions register as defective. Strikingly, substitution of R at posi- ity that the residual mutant protein is functionally impaired, we tion 225 causes RTT and prevents recruitment, whereas looked for other defects that might account for the RTT-like substitution of A at the same site does not cause RTT and phenotype. While many of the effects of MeCP2 mutations only recruits as efﬁciently as WT. become apparent when studied in the context of the whole or- Failure to efﬁciently recruit the cognate co-repressor to ganism, some properties, such as recruitment of the NCoR/ methylated sites in the genome suggested that P225R and P322L SMRT co-repressor complex to chromatin, can be tested using a might also be functionally defective in DNA methylation- simple transfection assay. P225R and P322L point mutations dependent transcriptional repression. To test this, untagged were introduced into a WT mouse Mecp2 cDNA fused to the C- MeCP2 was expressed in mouse tail ﬁbroblasts that were com- terminus of enhanced green ﬂuorescent protein (EGFP). promised in their ability to repress methylated reporter genes Interestingly, the Exome Aggregation Consortium (ExAC) data- due to deletion of the genes for Mecp2 and Mbd2 as previously base identiﬁes a hemizygous missense population variant also described (6). The co-transfected reporter construct encoding at P225 (P225A) that presumably does not cause RTT. We in- ﬁreﬂy luciferase was either unmethylated or methylated at ev- cluded this variant in our analysis and used Rett missense ery CpG (Fig. 4e) and repression was expressed as the ratio of lu- mutations R306C and R111G as negative controls for NCoR and ciferase activities expressed by unmethylated versus chromatin binding, respectively (Table 1). When WT EGFP- methylated constructs. Equivalent expression levels of WT, MeCP2 was expressed in mouse NIH-3T3 cells, green ﬂuores- R306C and P225R MeCP2 constructs were conﬁrmed by western cence co-localized with ‘DAPI bright spots’, which are blotting (Supplementary Material, Fig. S8a and b). Levels of Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 Human Molecular Genetics, 2018, Vol. 27, No. 14 | 2535 (a)(b) WT P322L null Stop MeCP2 H3 ** WT P322L (c) (d) WT P225R null flox MeCP2 H3 WT P225R (e) ES cells EBs -RA EBs +RA NPs neurons day 0 4 8 15 -LIF +RA plating harvest (f) (g) P225R P322L WT+lox WT 160 ns AABB AB pools MeCP2 100 NeuN H3 *** Figure 3. P322L and P225R mutations result in decreased MeCP2 levels in mouse brain and in vitro differentiated neurons. (a) Western blot of whole brain lysates from 6-week-old P322L and WT littermate males. Mecp2-null and Mecp2-stop samples are shown for comparison. (b) Quantiﬁcation of (a): the level of MeCP2 (MeCP2 signal/ H3 signal) as a percentage of mean WT is shown as mean (black line) and individual biological replicates (n¼3 animals). P¼0.0010 (**), two-tailed t-test with Welch’s cor- rection for unequal variances. (c) Western blot of whole brain lysates from 9-week-old P225R and WT littermate males. Mecp2-null and Mecp2-ﬂox samples are shown for comparison. (d) Quantiﬁcation of data shown in (c): the level of MeCP2 (MeCP2 signal/H3 signal) as a percentage of mean WT is shown as mean (black line) and indi- vidual biological replicates (n¼3 animals). P¼0.0128 (*), two-tailed unpaired t-test with Welch’s correction for unequal variances. (e) Neuronal differentiation protocol used to produce MeCP2-mutant neurons from mouse ES cells. EB: embryoid body, RA: retinoic acid, NP: neuronal precursor. (f) Representative western blot showing lev- els of MeCP2 protein present in neurons on day 15 of the differentiation scheme. Lanes A and B for each genotype are derived from independently targeted ES cell clones and WT pools are independent differentiations of the parental ES cell line. Histone H3 was used as a loading control for number of cells and NeuN as a control for equal neuronal differentiation. (g) Quantiﬁcation of MeCP2 levels in in vitro differentiated neurons. Two independent clones were differentiated two or three times for each genotype. MeCP2 level (MeCP2 signal/H3 signal) is normalized to the mean WTþloxP value. Data shown as mean (black line) and individual values. n¼5 (WT þloxP, P225R, P322L), n¼3 (WT pool) independent differentiations. Comparison to WTþloxP (two-tailed unpaired t-test with Welch’s correction for unequal variances): P225R P¼0.0416 (*), P322L P¼0.0023 (***) and WT (pool) P¼0.4868 (ns). P322L, however, were consistently low (30% of WT MeCP2; that the residual repression seen with R306C is due to recruit- data not shown), suggesting that lack of an EGFP tag rendered ment of another co-repressor, such as mSin3a, or to direct inter- the protein susceptible to degradation, as seen in neurons. As ference of dense methyl-CpG bound protein with the the amount of MeCP2 directly correlates with the strength of re- transcriptional machinery. pression in this assay, we excluded the P322L mutant from the transcriptional repression experiment. WT MeCP2 repressed the C-terminal deletions do not affect MeCP2 function or methylated luciferase efﬁciently, whereas repression by P225R recruitment was reduced by about half. The ability of R306C to repress was about 25% that of WT MeCP2 in this assay (Fig. 4f). These ﬁnd- Deletion-frameshift mutations in the C-terminal domain of ings correspond well with the behavior of the mutants in the MECP2, generally between nucleotides c.1100 and c.1200 (human TBL1X recruitment assay, where R306C was more severely af- e2 isoform numbering), are responsible for about 10% of all RTT fected than P225R (and P322L), supporting the view that failure cases. We chose to model two of the most common conﬁrmed to recruit the NCoR co-repressor underlies DNA methylation- Rett-causing CTDs in mice: c.1157–1197 D41 (CTD1) and c.1164– dependent transcriptional repression by MeCP2. It is possible 1207 D44 (CTD2; see Table 1). Native mouse MeCP2 contains a two WT+loxP P225R P322L WT pool MeCP2 (% WT+loxP) % WT mean % WT mean Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 2536 | Human Molecular Genetics, 2018, Vol. 27, No. 14 (a) WT R111G R306C P225A P225R P322L DAPI EGFP merge 10μm (b) EGFP WT R306C P225R P322L in IP in IP in IP in IP in IP NCoR1 300kD TBL1X 50kD 50kD HDAC3 (c) (d) EGFP mCherry DAPI merge ns WT ** R306C ** (e)(f) SV E 80 Luciferase ORF pA SV **** CpG **** **** Figure 4. P225R and P322L show impaired NCoR/SMRT recruitment to methylated DNA. (a) WT and mutant EGFP-MeCP2 fusions expressed in mouse NIH-3T3 cells. (b) Immunoprecipitation of EGFP-MeCP2 from transfected HEK 293 cells with GFP-Trap beads. NCoR/SMRT components NCoR1, TBL1X and HDAC3 detected in western blots of input and immunoprecipitated samples. (c) TBL1X recruitment assay. Co-transfection of EGFP-MeCP2 and TBL1X-mCherry expression constructs into NIH-3T3 cells. WT EGFP-MeCP2, but not R306C, recruits TBL1X-mCherry to heterochromatic foci. Scale bar 10 lm. (d) Quantiﬁcation of TBL1X recruitment efﬁciency of MeCP2 mutants. Percent of doubly transfected nuclei with TBL1X-mCherry/EGFP-MeCP2 spots from each independent transfection is shown as mean (black line) and individ- ual values. n¼7 (WT), 4 (R306C, P322L), 3 (P225A, P225R) independent transfections. MeCP2 mutants were compared to WT using a two-tailed Mann–Whitney test. R306C P¼0.0061 (**), P225A P¼0.1167 (ns), P225R P¼0.0167 (*), P322L P¼0.0061 (**). (e) Diagram of ﬁreﬂy luciferase expression construct pGL2-control consisting of an SV40 promoter, Luc coding sequence and SV40 polyadenylation signal and enhancer. The positions of CpG dinucleotides methylated by SssI.methylase are shown as black /y / lines. (f) Methylation-dependent transcriptional repression assay. Immortalized Mecp2 , Mbd2 mouse tail ﬁbroblasts transfected with SssI-methylated or unmethy- lated pGL2-control plasmid and an MeCP2 expression construct. Methylation-dependent repression expressed as the ratio of unmethylated/methylated (U/M) lucifer- ase activity and normalized to WT for each transfection. For each genotype the mean (black line) and individual values for four (R306C) or six (no MeCP2, WT, P225R) independent transfection experiments are shown. Each genotype compared to the WT value (100%) using a two-tailed one-sample t-test. No MeCP2 P<0.0001 (****), R306C P<0.0001 (****), P225R P¼0.0002 (****). WT R306C P225A P225R P322L WT R306C P225R U/M luc activity (%WT) % cells TBL1X recruited Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 Human Molecular Genetics, 2018, Vol. 27, No. 14 | 2537 amino acid deletion relative to the human protein (Fig. 5a)that mRNA levels in whole brain were not signiﬁcantly different coincides with the deletion hotspot in Rett patients. We modeled from WT littermate controls (Fig. 6a and b; Supplementary these Rett mutations in the mouse genome by restoring the two Material, Fig. S10e). Strikingly, CTD2 male hemizygous mice and human-speciﬁc amino acids and by adding the patient missense WT littermate controls were phenotypically indistinguishable tail (CTD1: HQPPX, CTD2: X) (Fig. 5a). We ﬁrst tested CTD1 and over the course of 1 year (Fig. 6e and f). Survival was 100% and CTD2 for methylated DNA-binding and NCoR/SMRT binding us- CTD2 mice behaved as WT in the elevated plus maze (Fig. 6g), ing transfection assays. Both were able to bind to methylated het- open ﬁeld test (Fig. 6h; Supplementary Material, Fig. S10f) and erochromatic foci in transfected NIH-3T3 cells (Fig. 5b) and to co- accelerating rotarod (Fig. 6i). In the wire suspension test CTD2 immunoprecipitate NCoR/SMRT components NCoR1, TBL1X and mice performed signiﬁcantly better than WT littermates HDAC3 (Fig. 5c) and mSin3a (Supplementary Material, Fig. S6b). (Supplementary Material, Fig. S10g). We conclude that the CTD2 Moreover, unlike P225R and P322L, both CTD1 and CTD2 were C-terminal truncation of MeCP2 has no overt phenotypic conse- able to recruit TBL1X-mCherry to methylated heterochromatic quences in this mouse model and therefore does not identify a foci as efﬁciently as WT MeCP2 (Fig. 5d). Thus, neither of the trun- previously unknown functional domain. cated proteins showed any deﬁciencies in MBD or NID function in these assays. ‘Humanization’ of the CTD2 allele reduces protein levels in neurons CTD1 and CTD2 mutant phenotypes reﬂect MeCP2 All Rett mutations previously studied in mice have replicated protein level the disease seen in patients, including the relative severity of The absence of overt defects in these assays raised the possibil- symptoms (5,6,23,24,31–33). Why should CTD2 be exceptional? ity that CTD1 and CTD2 disrupt other functional domains. To A potential explanation was that subtle differences in amino explore this possibility, we used CTD1 and CTD2 targeting vec- acid sequence between human and mouse MeCP2 proteins tors that retained all sequences downstream of the deletions to were involved. Despite 95% identity of MeCP2 protein between CTD1/y CTD2/y create Mecp2 and Mecp2 ES cell lines (Supplementary the two species, several amino acid differences persisted near Material, Fig. S9a and b). After Cre-mediated deletion of the the truncated region (Fig. 7a). To test whether these differences ﬂoxed Neo-Stop selection cassette, the mutated Mecp2 alleles altered stability of the mutant, we created a ‘humanized’ CTD2 differed from WT only at the mutation site and a loxP site up- allele (CTD2hu) in mouse ES cells by making three single nucle- stream of exon 3. When these lines were differentiated into otide alterations: T376P, M380V and S387P (Fig. 7a). These neurons in vitro, both genotypes produced truncated MeCP2 pro- changes ensure that the amino acid sequence of the C-terminus tein as expected, but surprisingly the amount of protein differed of MeCP2 in patient and mouse alleles is identical from E298, greatly. CTD1 retained only 3% the WT level, whereas the which is N-terminal to the NID, to the stop codon. CTD2hu ES amount of CTD2 MeCP2 was not signiﬁcantly different from WT cell clones were produced (Supplementary Material, Fig. S11a controls (Fig. 5e and f). At the transcript level, CTD1 had only and b) and differentiated into neurons for comparison with 34% WT mRNA whereas CTD2, like P225R and P322L, expressed CTD1, CTD2 and WT þ lox controls. Western blots showed a WT-levels of mRNA (Supplementary Material, Fig. S9c). dramatic destabilization of CTD2hu, which was now indistin- The extremely low level of MeCP2 protein in CTD1 neurons guishable from CTD1 (7.4 and 5.0% WT þ lox, respectively, predicted that mice would show a severe phenotype, similar to Fig. 7b and c). The original CTD2 MeCP2 protein remained essen- P322L and Mecp2-null mice. To test this, we produced CTD1 mice tially identical to WT þ lox controls (93%). CTD2hu neurons also using CRISPR/Cas9 cutting and repair with an oligonucleotide contained less mRNA than WT þ lox and CTD2, again resem- template in mouse zygotes. After conﬁrming the presence of the bling CTD1 (Supplementary Material, Fig. S11c). Thus, altering mutation by sequencing (data not shown) and southern blot the sequence context of the non-pathogenic CTD2 mouse allele (Supplementary Material, Fig. S10a and b), analysis of CTD1 to make it resemble the patient mutation more closely resulted founder males showed that the amount of MeCP2 in whole brain in a drastic reduction in the amount of MeCP2 in neurons. was reduced to about 10% of that in WT controls (Fig. 6a and b) To further conﬁrm our hypothesis that CTD2 MeCP2 is unsta- and the mRNA to 45% of WT (Supplementary Material, Fig. S10e). CTD2 ble in human neurons we created a knock-in MECP2 allele in A cohort of CTD1 male mice was scored for Rett-like phenotypes, Lund human mesencephalic cells (LUHMES) cells, a human neuro- which progressed rapidly (Fig. 6c), giving a median survival of 20 nal progenitor cell line derived from human embryonic ventral weeks (Fig. 6d). The weights of CTD1 mice were not signiﬁcantly mesencephalic tissue which can be differentiated in vitro to pro- different to those of WT littermates (Supplementary Material, Fig. duce a uniform population of mature dopaminergic neurons S10c). This cohort was still on a mixed CBA/C57BL6 genetic back- CTD2 (34,35). Knock-in clones containing MECP2 alleles on the active ground whereas small size is usually seen in Mecp2-mutant males X chromosome, and thus expressing MeCP2 CTD2 protein, were on a more inbred C57BL6 background. produced using CRISPR/Cas9 cutting and an oligodeoxynucleotide The instability of CTD1 made it impossible to deduce repair template as previously described (35)(Supplementary whether the C-terminal truncation mutation gave rise to a novel Material,Fig.S11dand e). After differentiating for 9 days to mature functional defect in MeCP2 as at 10% of WT littermate controls neurons, the level of CTD2 MeCP2 protein was severely reduced the level of protein was similar to the level of full-length protein Stop/y compared to the WT parental cells and unmodiﬁed WT clones in Mecp2 mice, which resemble Mecp2-null mice (22). On the (Fig. 7d and e). MeCP2 protein levels in CTD2 clones ranged from 7 other hand, the fully stable CTD2 mutant protein offered the to 20% of the parental WT cells. These ﬁndings in mice, mouse ES chance to test the importance of this C-terminal region. CTD2 mice were therefore produced using standard injection of ES cell-derived neurons and human neurons strongly suggest that cells into mouse blastocysts and veriﬁed by southern blots RTT resulting from CTD mutations is due to severe MeCP2 deﬁ- (Supplementary Material, Fig. S10d) and sequencing (data not ciency rather than the loss of essential domains in the C-terminal shown). As in cultured neurons, CTD2 MeCP2 protein and region. Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 2538 | Human Molecular Genetics, 2018, Vol. 27, No. 14 (a) (b) (d) (c) (e) (f) Figure 5. C-terminal truncation mutants CTD1 and CTD2 show normal NCoR/SMRT recruitment but differ in neuronal protein levels. (a) Comparison of the human and mouse MeCP2 CTD regions for wild-type (WT) and the patient mutations CTD1 and CTD2. Amino acid changes and absent amino acids in the mouse protein are shown in red, amino acids added to mouse CTD1 and CTD2 proteins in blue and missense/nonsense changes arising due to deletion/frameshift in bold type. (b) WT, CTD1 and CTD2 EGFP-MeCP2 fusions expressed in mouse NIH-3T3 cells. (c) Immunoprecipitation of EGFP-MeCP2 from transfected HEK 293 cells with GFP-Trap beads. NCoR/ SMRT components NCoR1, TBL1X and HDAC3 detected in western blots of input and immunoprecipitated samples. (d) Quantiﬁcation of TBL1X recruitment efﬁciency of MeCP2 mutants. Percent of doubly transfected nuclei with TBL1X-mCherry/EGFP-MeCP2 spots in each independent transfection shown as mean (black line) and in- dividual values. WT n¼7, R306C, CTD1, CTD2 n¼4 independent transfections. Genotypes were compared to WT using a two-tailed Mann–Whitney test. R306C P¼0.0061 (**), CTD1 P¼0.0727 (ns), CTD2 P¼0.6485 (ns). (e) Representative western blot showing levels of MeCP2 protein present in neurons on day 15 of differentiation (7 days af- ter plating). Lanes A and B for each genotype are derived from independently targeted ES cell clones and WT pools are independent differentiations of the parental ES cell line. An antibody against the N-terminus of MeCP2 detects both full-length and truncated (D) protein. (f) Quantiﬁcation of MeCP2 levels in in vitro differentiated neurons. Two independent clones were differentiated 2 or 3 times for each genotype. MeCP2 level (MeCP2 signal/H3 signal) is normalized to the mean WT þ loxP value. Data shown are mean (black line) and ﬁve (WTþloxP, CTD1) or three (CTD2, WT pools) independent differentiations. Comparison to WTþloxP (two-tailed unpaired t-test with Welch’s correction for unequal variances): CTD1 P¼0.0021 (***), CTD2 P¼0.1874 (ns) and WT (pool) P¼0.4868 (ns). Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 Human Molecular Genetics, 2018, Vol. 27, No. 14 | 2539 ns (a)(b) 200 WT CTD1 CTD2 null MeCP2 MeCP2Δ H3 *** WT CTD1 CTD2 (c)(d) 10 100 CTD1 WT CTD1 8 80 null WT 6 60 null 4 40 2 20 0 0 4 8 12 16 20 24 28 32 0 4 8 12 16 20 24 28 32 36 Age (weeks) Weeks (e)(f) 60 CTD2 WT CTD2 WT 0 0 4 8 12 16 20 24 28 32 36 40 44 48 52 4 8 12 16 20 24 28 32 36 40 44 48 52 Age (weeks) Age (weeks) (g)(h)(i) Elevated Plus Maze Open Field (centre) Rotarod ns 100 ns 0.10 200 ns 80 0.08 0.06 ns 40 0.04 0.02 0 0.00 0 Closed Open CTD2 WT 123 Day CTD2 WT Figure 6. CTD1 and CTD2 mice show contrasting phenotypes due to different levels of MeCP2 in brain. (a) Western blot of whole brain lysates from 6-week-old CTD1, CTD2 and WT males. A Mecp2-null sample is shown for comparison. CTD1 and CTD2 mice express a truncated protein, MeCP2D.(b) Quantiﬁcation of (a): the level of MeCP2 (MeCP2 signal/H3 signal) as a percentage of mean WT shown as mean (black line) and individual values for each genotype (n¼3 animals). Comparison to WT (two-tailed unpaired t-test with Welch’s correction for unequal variances): CTD1 P¼0.0008 (***) and CTD2 P¼ 0.1773 (ns). (c) CTD1 hemizygous males display RTT-like phenotypes. Mean aggregate phenotypic score6SD plotted for mutants and WT male littermates, CTD1 n¼11, WT n¼7 (at 4 weeks). (d) Kaplan–Meier survival plot for CTD1 hemizygous males and WT littermates shown in (c). (e) Phenotypic scoring of CTD2 males and WT littermates. Mean aggregate phenotypic score6SD plotted for mutants and WT male littermates, CTD2 n¼16, WT n¼17. (f) Mean weight6SD for animals shown in (e). (g) Elevated plus maze. Percentage of time spent in the closed and open arms is shown as mean (black line) and individual values. Percentage time in open and closed arms for each genotype was compared using two-tailed un- paired t-tests: closed arms P¼0.2238 (ns), open arms P¼0.3713 (ns). (h) The ratio of distance travelled in the central zone of the open ﬁeld arena to total distance trav- elled shown as mean (black line) and individual values. Two-tailed unpaired t-test test P¼0.9394 (ns). (i) Accelerating rotarod test. The latency to fall is shown as mean (black line) and individual data points on each day. Two-way ANOVA (repeated measures) [genotype effect, F(1, 17)¼9.45110 , P¼ 0.9976 (ns)]. (g)–(i) CTD2 n¼9, WT n¼10. causal missense mutations disrupt the interactions with meth- Discussion ylated DNA or with the NCoR/SMRT co-repressor complexes Biochemical and cell biological studies of MeCP2 have identiﬁed (11). We were intrigued by a number of RTT mutations that numerous interacting proteins that potentially mediate its leave the MBD and the NID intact, and by the possibility that function. So far, only two of these are implicated in RTT, as they might direct us to additional functional domains that had Aggregate Score % time in arms Aggregate Score Distance: centre/total Weight (g) Percent survival Latency to fall (s) % WT mean Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 2540 | Human Molecular Genetics, 2018, Vol. 27, No. 14 (a) (b)(c) (d) (e) Figure 7. A humanized CTD2 allele expresses low levels of truncated MeCP2. (a) The amino acid sequences of the extreme C-termini of CTD1, CTD2 and CTD2hu mouse alleles and their corresponding human RTT alleles. Missense/nonsense changes caused by the deletion-frameshift are in bold type. Residues where the mouse allele differs from human are shown in red. Changes between CTD2 and CTD2hu are underlined. (b) Western blot showing levels of MeCP2 protein present in neurons on day 15 of differentiation (7 days after plating). Lanes A, B (and C) for each genotype are derived from independent mouse ES cell clones. An antibody against the N-terminus of MeCP2 detects both full-length and truncated (D) protein. (c) Quantiﬁcation of MeCP2 levels in in vitro differentiated neurons. Two or three independent clones were differentiated for each genotype. MeCP2 level (MeCP2 signal/H3 signal) is normalized to the mean WTþloxP value. Data shown are mean (black line) and individual val- ues for three (WTþloxP, CTD2hu) or two (CTD1, CTD2) independent differentiations. Comparison to WTþloxP (two-tailed unpaired t-test with Welch’s correction for unequal variances): CTD1 P¼0.0108 (*), CTD2 P¼0.5450 (ns) and CTD2hu P¼0.0111 (*). (d) Western blot showing levels of MeCP2 protein present in LUHMES-derived hu- man neurons on day 9 after precursor plating. Differentiations of one MECP2-null clone, two WT pools, two unmodiﬁed WT clones and three CTD2 LUHMES clones are shown. An antibody against the N-terminus of MeCP2 detects both full-length and truncated (D) protein. (e) Quantiﬁcation of MeCP2 levels in LUHMES-derived neurons. Three independent differentiations were performed for each of the clones shown in (d). MeCP2 level (MeCP2 signal/H3 signal) is normalized to the WT pool value for each differentiation. Data shown are mean (black line) and individual values for three independent differentiations. Comparison to WT A (two-tailed unpaired t-test): WT B P¼0.0798 (ns), CTD2 A P¼0.0015 (***), CTD2 B P¼0.0039 (***), CTD2 C P¼0.0027 (***). previously escaped detection. To be certain that the mutations detected several times in classical RTT (7), but has subsequently are genuinely causative of the disorder, we conﬁned our atten- been classiﬁed as a relatively common and phenotypically be- tion to variants shown to be absent in parental DNA. The impor- nign population variant (7,8). By studying C-terminal trunca- tance of this criterion is illustrated by E397K, which has been tions and the missense mutations P225R and P322L, which are Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 Human Molecular Genetics, 2018, Vol. 27, No. 14 | 2541 recurrent RTT mutations remote from the two known interac- and nonsense mutations R168X (23) and R255X (32). Our results tion sites, we failed to establish new functional domains of show that MeCP2 deﬁciency due to protein destabilization alone MeCP2. Instead, our results can provide coherent molecular is a major underlying cause of RTT. Although we only tested explanations for the deleterious effects of these mutations by two CTD mutations, we suggest that severe MeCP2 destabiliza- considering MeCP2 as a recruiter of histone deacetylase com- tion is a mechanism that applies generally to this RTT mutation plexes to chromatin (29). Further evidence that it is the confor- category. Why mutations outside the key functional domains mation and/or stability of MeCP2 that is affected by these should confer instability is unknown. With the exception of the mutations rather than functional domains containing P225, MBD and NID domains, whose 3D conformation has been deter- P322 or the CTD region comes from a severely truncated mouse mined (12,13), most of the protein is thought to be unstructured knock-in allele, DNIC, which despite lacking all of these sites (40). Surprisingly, the instability of CTD2 in the context of hu- shows a very mild phenotype, distinct from that seen in RTT man MeCP2 was suppressed by changing three nearby amino mouse models (14). acids to match the mouse MeCP2 sequence. This serendipi- It is clear from studies of patient severity (18,33,36,37) and tously allowed us to determine that the deletion of the MeCP2 mouse models (5,6,24,31) that mutations in the MBD and the CTD beyond amino acid 388, thereby losing 100 amino acids NID, such as T158M and R306C, are somewhat less severe on av- from the C-terminus, has no deleterious phenotypic consequen- erage than complete absence of the protein. This suggests that ces. We conclude therefore that this mutation does not identify these mutant proteins retain some function, which could come a novel functional domain, but causes RTT in humans purely from a number of sources. T158M does not completely abolish due to drastic destabilization of the protein. binding to methylated DNA (24) and some residual binding of The dramatic difference in the stability of CTD2 conferred by R306C to NCoR components can be seen in this study (Fig. 4b; human versus mouse sequences indicates that subtle changes Supplementary Material, Fig. S6b). All MeCP2 proteins tested in protein primary structure can have dramatic consequences could bind to mSin3a (Supplementary Material, Fig. S6b). It has for MeCP2 abundance. It is also interesting that CTD alleles that been proposed that other domains such as AT-hooks (38,39) result in reduced protein levels also have less mRNA, although may also contribute low-level function of MeCP2 when the MBD the decrease is less severe. Further investigation of the precise and NID are mutated. cause of CTD instability will be a focus of future research. P225R leads to moderate destabilization of MeCP2, but it also Importantly, the ﬁnding that CTD mutations do not appear to interferes with its ability to recruit the co-repressor subunit affect MeCP2 function may offer therapeutic opportunities. A re- TBL1X to chromatin and to impose DNA methylation- cent study has demonstrated the use of proteasome inhibitors dependent transcriptional repression. Although P225 is in an to increase the level of unstable T158M MeCP2 (46). unstructured region of the protein, it has been reported that Alternatively, small molecules that bind to the mutant CTD and conformational changes occur on binding to DNA (40). P225R affect its conformation may improve stability and therefore pro- may hinder the ability of MeCP2 to form the required bridge be- vide clinical beneﬁt. tween chromatin and co-repressor complex. It is also possible that the same conformational change leads to instability of the Materials and Methods protein. The combination of these two partial defects in stability and function can explain the moderate RTT-like phenotype of Mecp2 mutant mouse alleles these mice. Importantly, a rare human population variant at the RTT mutations P225R, P322L, CTD1, CTD2 and CTD2hu were in- same amino acid, P225A, does not interfere with co-repressor troduced into a targeting vector (20)(Fig. 2a) containing WT recruitment. The strong correlation between the ability to re- Mecp2 sequences and a Cre-excisable selection cassette using cruit co-repressor to methylated DNA and phenotypes in mice the QuikChangeII XL Site-Directed Mutagenesis kit (Agilent and humans strengthens the notion that this is a key role of Technologies). Targeting vectors were linearized at the 3 end MeCP2. using NotI and electroporated into 129/Ola E14Tg2a mouse ES The other non-canonical missense mutation, P322L, severely cells (a gift from A. Smith, University of Edinburgh). Correctly destabilizes MeCP2 and this alone can account for the resulting targeted clones were identiﬁed by an initial PCR screen: severe phenotype. The C-terminal truncations CTD1 and CTD2hu also severely destabilize the protein in neurons, but do 0 0 forward primer: 5 -TCACCATAACCAGCCTGCTCGC-3 not detectably compromise several aspects of its function in cell 0 0 reverse primer: 5 -ATTCGATGACCTCGAGGATCCG-3 ) transfection-based assays. Emphasizing this point, CTD2 mice, followed by Southern blotting of candidate clones and sequenc- which lack 100 amino acids from the C-terminal end of MeCP2, ing of the mutation sites (Supplementary Material, Figs S1 and are viable, fertile and phenotypically normal in a series of be- S9). A number of clones that had recombined to include the havioral assays. We note that this truncation removes an ﬂoxed selection cassette but no RTT mutation were also veriﬁed activity-dependent phosphorylation site at S421 (S423 in human for use as controls (WT þ lox). MeCP2) that has previously been implicated in neuronal func- tion, including dendritic patterning and spine morphogenesis (41–44). The region has also been implicated in regulation of Differentiation of ES cells into neurons micro-RNA processing and dendritic growth (45). Although loss of these functional domains evidently does not make a detect- ES clones were transiently transfected with a pCAGGS-Cre ex- able contribution to the major RTT-like phenotypes in mice, we pression plasmid and clones that had deleted the Neo selection cannot exclude the possibility that subtle phenotypic conse- cassette were veriﬁed by PCR, Southern blotting and sequencing quences were not detected by our assays or that they are impor- before differentiation into neurons. Two independently targeted tant for higher functions present in humans but not in mice. clones were used for each genotype, and each clone was differ- MeCP2 instability has previously been noted as a result of entiated into neurons two or three times. Differentiation was mutations that also disrupt the functional domains of MeCP2, carried out using a 4-/4þ retinoic acid (RA) procedure as previ- including missense mutations T158A (31), T158M and R133C (24) ously described (47,48). Neural precursor cells were seeded onto Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 2542 | Human Molecular Genetics, 2018, Vol. 27, No. 14 5 2 6 cm dishes at a density of 1.510 cells/cm and harvested after TGCCACTGCTCCCACCCCACCAGCCCCCCTGAGCCTCAGGACTTG 7 days by scraping into phosphate-buffered saline (PBS), pellet- AGCAGCAGCATCTGCAAAGAAGAGAAGATGC) (Sigma) into ing the cells and snap freezing. C57BL6/J: CBA/CaOlaHsd F2 fertilized oocytes as previously de- scribed (51). Correctly mutated founder animals were identiﬁed by sequencing across the deletion site. One CTD1 hemizygous Generation of CTD2 LUHMES cell lines male founder was able to breed and female heterozygous off- The LUHMES cell line (ATTC CRL-2927) was obtained from ATCC spring were checked by sequencing and Southern blot and cultured according to the methods described in Scholz et al. (Supplementary Material, Fig. S10a and b) and used for further (49), with some minor alterations. All vessels were coated in breeding of the line. poly-L-ornithine and ﬁbronectin overnight at 37 C. Proliferating LUHMES cells were seeded at 210 cells/T75 ﬂask every 2 days. Phenotypic analysis For differentiation, 2.510 cells were seeded in a T75 ﬂask for the ﬁrst 2 days of the protocol and on day 2 cells were seeded at Cohorts of hemizygous mutant male mice and WT littermates 610 cells/10 cm dish. During differentiation a half-media were weighed and scored weekly for a range of RTT-like pheno- change was performed on day 6 and neurons were harvested types to give an aggregate score between 0 and 12 as previously for protein on day 9. described (20,25). Cohorts of at least eight mutants and eight To introduce a CTD2 knock-in mutation, LUHMES cells were WT littermates were scored for each genotype, with higher transfected with plasmid pSpCas9(BB)-2A-GFP (pX458) (a gift numbers being used where available to allow for any losses from Feng Zhang, Addgene plasmid #48138) containing the unrelated to the mutation, such as ﬁghting. Scoring was carried 0 0 sgRNA sequence 5 -TCCTCGGAGCTCTCGGGCTC-3 , and single- out blind to genotype and to previous scores. Animals that stranded oligodeoxynucleotide 5 CCATCACCACCACTCAGAGTC scored a maximum score of 2 for tremor, breathing or general CCCAAAGGCCCCCGTGCCACTGCTCCCACCCCTGCCGCCCTGAGC condition or which had lost 20% of their body weight had CCCAGGACTTGAGCAGCAGCGTCTGCAAAGAGGAGAAGATGCCC reached the severity limit of the experiment according to the AGAGGAGGCT3 (Sigma, desalted). Cells were transfected by Home Ofﬁce license and were humanely culled. These animals Nucleofection (Lonza) using a Basic Nucleofector kit for primary were counted as having ‘died’ for the purposes of survival data. neurons (VAPI-1003) and a Nucleofector II device, and ﬂuores- Animals of any genotype which were culled for reasons not cence activated cell sorting (FACS)-sorted to isolate clones as linked to the mutation, such as ﬁghting with cage mates, were previously described (35). removed from survival plots at that point (censored data). LUHMES clones were analyzed by sequencing of genomic P225R, P322L and CTD2 animals had been backcrossed to DNA and cDNA and by Southern blotting (Supplementary C57BL6/J for three generations and CTD1 animals only once. Material, Fig. S11d and e) to isolate clones that were either ho- Behavioral testing was carried out on cohorts of mice that mozygous for the CTD2 mutation, or heterozygous and express- had been backcrossed to C57BL6/J for four generations. Testing ing the CTD2 mutation from the active X-chromosome took place over a 9-day period in the order: day 1: elevated plus (LUHMES cells are female, XX). Three CTD2 clones (two homozy- maze (115 min trial), day 2: open ﬁeld test (120 min trial), day gous and one heterozygous) and two unmodiﬁed WT control 3: wire suspension test (3 trials of 30 sec separated by 15 min), clones were used for further analysis. day 6: accelerating rotarod habituation (5 min at 4 rpm), days 7– LUHMES clones were differentiated to mature neurons as 9: accelerating rotarod trials (4 trials per day separated by 1 h). previously described (49). A null clone (H4) was described previ- Tests were performed as previously described (24,25). ously (35). All clones and the parental cell line underwent three Hemizygous male mutant mice and WT littermates were tested independent differentiations and were harvested on day 9 of at an age appropriate to the development of symptoms for that differentiation by scraping into PBS, pelleting the cells and snap line: testing was started for P322L at 6 weeks, P225R 11 weeks freezing. and CTD2 18 weeks. A cohort size of 10 animals per genotype was chosen as the largest number of animals that could reason- Establishment of Mecp2 mutant mouse lines ably be tested in a single session. Mice were housed in mixed mutant and WT littermate groups. Mice were tested blind to ge- P225R, P322L and CTD2 mouse lines were generated by injection notype and in order of ID number, giving a random order of mu- of ES cells into mouse blastocysts using standard methods. tant and WT animals. Positions on the rotarod were assigned to Chimeric males were crossed with CMVCre deleter females (50) ensure equal representation of mutant and control animals at to remove the selection cassette. This and subsequent genera- each of the ﬁve positions. tions were bred by crossing Mecp2-mutant heterozygote females with C57BL/6J WT males. The genotypes of mutant lines were conﬁrmed by Southern blotting (Supplementary Material, Figs RNA preparation and qRT-PCR S2a and b and S10d) and sequencing. Routine genotyping was Total cellular RNA was prepared from in vitro differentiated neu- performed by PCR across the loxP ‘scar’ site remaining in the rons (7 days after plating, 10 cells) and mouse brain (half brain) mutant alleles after selection cassette removal (forward primer: 0 0 using TriReagent (Sigma). cDNA was prepared using a p5 5 -TGGTAAAGACCCATGTGACCCAAG-3 , reverse primer: 0 0 QuantiTect kit (Qiagen) and ampliﬁed in a qPCR reaction using p7 5 -GGCTTGCCACATGACAAGAC-3 , WT 416bp, WT þ loxP SensiMix SYBR and Fluoroscein Master Mix (Bioline) using pri- 558bp). 0 0 mers for Mecp2 (forward: 5 -ACCTTGCCTGAAGGTTGGAC-3 , re- The CTD1 mouse line was created by pronuclear injection of 0 0 Cas9 protein (Integrated DNA Technology), synthetic tracrRNA verse: 5 -GCAATCAATTCTACTTTAGAGCGAAAA-3 ) and control 0 0 and crRNA (target sequence ACCTGAGCCTGAGAGCTCTG) Cyclophilin A (forward: 5 -TCGAGCTCTGAGCACTGGAG-3 , re- 0 0 verse: 5 -CATTATGGCGTGTAAAGTCACCA-3 ). Samples were (Sigma) and an oligonucleotide repair template (GCACCATCATC ACCACCATCACTCAGAGTCCACAAAGGCCCCCA run in triplicate and the amount of Mecp2 cDNA calculated for Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 Human Molecular Genetics, 2018, Vol. 27, No. 14 | 2543 each biological replicate using the ‘double delta’ method after double-transfected cells with DAPI/EGFP/mCherry co-localized correction of C values using a standard curve. spots was then calculated for each genotype in each indepen- dent transfection. Protein extracts and western blotting In vitro differentiated ES cell-derived and LUHMES-derived neu- Co-immunoprecipitation assay rons and whole brain samples were prepared for western blot- EGFP-MeCP2 expression constructs were transfected into HEK ting as previously described (3). Extracts were run on TGX 4–20% 293FT cells (R70007, ThermoFisher) using Lipofectamine and gradient gels (BioRad) loading extract equivalent to 5 10 nu- cell pellets were harvested 24 h after transfection and snap fro- clei per well. Samples were run on duplicate gels and trans- zen. Cell pellets were treated with Benzonase (Sigma) in low salt ferred to nitrocellulose membrane overnight in the cold at 25 V. NE1 buffer (20 mM HEPES pH7.5, 10 mM NaCl, 1 mM MgCl 0.1% 2, Western blots were processed as described previously (3) using Triton-X100, 10 mM b-mercaptoethanol and protease inhibitors) the following antibodies: anti-MeCP2 (N-terminus): mouse to release chromatin bound proteins and then extracted with monoclonal Men-8 (Sigma), anti-NeuN: rabbit polyclonal ABN78 150 mM NaCl. Supernatant extracts were mixed with GFP- (Millipore) and anti-histone H3: rabbit polyclonal ab1791 Trap _A beads (Chromotek) and after washing beads were (Abcam). Western blots were developed with IR-dye secondary boiled in SDS-PAGE sample buffer and released proteins and in- antibodies (IRDye 800CW donkey anti-mouse, IRDye 680LT don- put extract samples run on 4–15% TGX gradient gels (BioRad) key anti-rabbit, LI-COR Biosciences) and scanned using a LI-COR and blotted as above. Co-immunoprecipitated proteins were Odyssey machine. Images were quantiﬁed using Image Studio detected with the following antibodies: anti-NCoR: rabbit poly- Lite software (LI-COR Biosciences). The ratio of NeuN: histone H3 signals for each lane was used to check for equal neuronal clonal A301–145A (Bethyl Laboratories), anti-TBL1X: rabbit poly- in vitro differentiation and clones that did not differentiate well clonal ab24548 (Abcam), anti-HDAC3: mouse monoclonal 3E11 were discarded. MeCP2 levels were normalized to the histone (Sigma), anti-mSin3a: rabbit polyclonal ab3479 (Abcam) and H3 signal for each lane to compare the amount of MeCP2/nu- anti-EGFP: mouse monoclonal Living Colors JL-8 (Clontech). cleus between samples. Methylation-dependent repression assay EGFP-tagged cDNA constructs and transfection assays WT and mutant mouse Mecp2 e1 cDNAs were cloned into the Mouse Mecp2 cDNA (e2 isoform) was cloned into the vector expression vector pcDNA3.1(þ) IRES GFP, a gift from Kathleen L pEGFP-C1 (Clontech) to create in-frame fusions with an N-termi- Collins, Addgene plasmid #51406 (52). The methylation- nal EGFP tag. Missense mutations were introduced using the dependent repression assay was performed as previously de- QuikChangeII XL Site-Directed Mutagenesis kit (Agilent scribed (6) by co-transfecting MeCP2 expression plasmid, Fireﬂy Technologies) and truncations by PCR amplifying the cDNA luciferase (Luc) construct pGL2-control (with or without SssI with a modiﬁed reverse primer before cloning into the expres- methylation) and Renilla luciferase (Ren) transfection control sion vector. A vector expressing mCherry-tagged mouse TBL1X /y / plasmid into a Mecp2 , Mbd2 mouse tail ﬁbroblast cell line was described previously (30). For localization of EGFP-MeCP2 (6) using Lipofectamine. Luciferase signal was measured using and co-localization of EGFP-MeCP2/TBL1X-mCherry constructs the Dual Luciferase Assay kit (Promega). The ratio Luc/Ren sig- were transfected into NIH-3T3 mouse ﬁbroblasts (93061524, nal was calculated for each well and the mean unmethylated: ECACC) growing on glass coverslips using Lipofectamine (Thermo Fisher). Cells were ﬁxed in 4% paraformaldehyde 48 h mean methylated (U/M) value calculated for each genotype. after transfection, stained with DAPI and mounted in ProLong Each combination was repeated in triplicate in each experiment Diamond mountant (Thermo Fisher). Images were captured us- and each MeCP2 genotype was independently transfected at ing a Leica SP5 confocal microscope with 63 objective. For least four times. To check for equal expression of MeCP2 pro- western blotting to test expression levels of EGFP-MeCP2 WT tein, total protein extract from triplicate wells transfected with and mutant proteins cells were transfected with EGFP-MeCP2 either unmethylated or methylated luciferase template and and TBL1X-mCherry expression constructs as above and har- each of the MeCP2 expression constructs was western blotted vested 48 h after transfection by trypsinization. Cell pellets and the level of MeCP2 (normalized to c-tubulin) determined. were prepared for western blotting as described above for Primary antibodies: anti-MeCP2 rabbit monoclonal D4F3 (Cell in vitro differentiated neurons. MeCP2 was detected using Signaling Technology) and anti-c-tubulin mouse monoclonal mouse monoclonal Men-8 (Sigma) and anti-c-tubulin mouse GTU-88 (Sigma). monoclonal GTU-88 (Sigma) was used as a loading control. For quantiﬁcation of TBL1X-mCherry co-localization with EGFP-MeCP2, 3–7 independent transfections were performed for Statistical Analysis each genotype, with each transfection experiment including All statistical analysis was performed using GraphPad Prism EGFP-MeCP2 WT (positive control) and EGFP-MeCP2 R306C (neg- (GraphPad Software). Datasets were compared using Student’s ative control). Fields of cells were selected for scoring based on t-tests (with or without Welch’s correction for unequal variance, the presence of an EGFP (MeCP2) signal in at least one cell in the as appropriate), apart from the TBL1X-mCherry co-localization ﬁeld, without observing the mCherry channel. Cells with both assay and wire suspension test that were analyzed using a non- EGFP and mCherry ﬂuorescence were then scored for the pres- parametric Mann–Whitney test due to the nature of the data. ence or absence of TBL1X-mCherry spots in the nucleus, co-lo- Rotarod data were tested using two-way repeated measures calized with EGFP-MeCP2 and the DAPI bright spots. In each ANOVA, followed by post-hoc testing for signiﬁcance on each transfection at least 15 cells on 4 coverslips (total 60 cells per ge- notype) were counted, blind to the genotype. The percentage of day using the Holm–Sidak method. Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 2544 | Human Molecular Genetics, 2018, Vol. 27, No. 14 Cummings, B.B. et al. (2016) Analysis of protein-coding ge- Study Approval netic variation in 60, 706 humans. Nature, 536, 285–291. All animal experiments were performed under a United 9. Ebert, D.H., Gabel, H.W., Robinson, N.D., Kastan, N.R., Hu, Kingdom Home Ofﬁce project license (PPL no. 60/4547). L.S., Cohen, S., Navarro, A.J., Lyst, M.J., Ekiert, R., Bird, A.P. et al. (2013) Activity-dependent phosphorylation of MeCP2 threonine 308 regulates interaction with NCoR. Nature, 499, Data Availability 341–345. Data from this work are available from the authors. 10. Heckman, L.D., Chahrour, M.H. and Zoghbi, H.Y. (2014) Rett-causing mutations reveal two domains critical for Supplementary Material MeCP2 function and for toxicity in MECP2 duplication syn- drome mice. Elife, 3, e02676. Supplementary Material is available at HMG online. 11. Lyst, M.J. and Bird, A. (2015) Rett syndrome: a complex disor- der with simple roots. Nat. Rev. Genet., 16, 261–275. 12. Ho, K.L., McNae, I.W., Schmiedeberg, L., Klose, R.J., Bird, A.P. Acknowledgements and Walkinshaw, M.D. (2008) MeCP2 binding to DNA We would like to thank Alan McClure for animal husbandry and depends upon hydration at methyl-CpG. Mol. Cell, 29, Matthew Lyst and Dirk-Jan Kleinjan for critical reading of the 525–531. manuscript. A.B. is a member of the Simons Initiative for the 13. Kruusvee, V., Lyst, M.J., Taylor, C., Tarnauskaite, Z., Bird, A.P. Developing Brain at the University of Edinburgh. and Cook, A.G. (2017) Structure of the MeCP2-TBLR1 complex reveals a molecular basis for Rett syndrome and related dis- Conﬂict of Interest statement. A.B. is a member of the Board of orders. Proc. Natl. Acad. Sci. U S A, 114, E3243–E3250. ArRETT, a company based in the USA with the goal of develop- 14. Tillotson, R., Selfridge, J., Koerner, M.V., Gadalla, K.K.E., Guy, ing therapies for Rett syndrome. J., De Sousa, D., Hector, R.D., Cobb, S.R. and Bird, A. (2017) Radically truncated MeCP2 rescues Rett syndrome-like neu- Funding rological defects. Nature, 550, 398–401. 15. Gabel, H.W., Kinde, B., Stroud, H., Gilbert, C.S., Harmin, D.A., This work was supported by Wellcome (Investigator Award Kastan, N.R., Hemberg, M., Ebert, D.H. and Greenberg, M.E. 107930, PhD Studentship 109090 to L.F.) and the Rett Syndrome (2015) Disruption of DNA-methylation-dependent long gene Research Trust. Funding to pay the Open Access publication repression in Rett syndrome. Nature, 522, 89–93. charges for this article was provided by Wellcome. 16. Kinde, B., Wu, D.Y., Greenberg, M.E. and Gabel, H.W. (2016) DNA methylation in the gene body inﬂuences References MeCP2-mediated gene repression. Proc. Natl. Acad. Sci. U S A, 1. Amir, R.E., Van den Veyver, I.B., Wan, M., Tran, C.Q., Francke, 113, 15114–15119. U. and Zoghbi, H.Y. (1999) Rett syndrome is caused by muta- 17. Lagger, S., Connelly, J.C., Schweikert, G., Webb, S., Selfridge, J., Ramsahoye, B.H., Yu, M., He, C., Sanguinetti, G., Sowers, tions in X-linked MECP2, encoding methyl-CpG-binding pro- L.C., Walkinshaw, M.D. and Bird, A. (2017) MeCP2 recognizes tein 2. Nat. Genet., 23, 185–188. 2. Lewis, J.D., Meehan, R.R., Henzel, W.J., Maurer-Fogy, I., cytosine methylated tri-nucleotide and di-nucleotide Jeppesen, P., Klein, F. and Bird, A. (1992) Puriﬁcation, se- sequences to tune transcription in the mammalian brain. quence, and cellular localization of a novel chromosomal PLoS Genet., 13, e1006793. protein that binds to methylated DNA. Cell, 69, 905–914. 18. Bebbington, A., Percy, A., Christodoulou, J., Ravine, D., Ho, G., Jacoby, P., Anderson, A., Pineda, M., Ben Zeev, B., Bahi- 3. Ross, P.D., Guy, J., Selfridge, J., Kamal, B., Bahey, N., Tanner, K.E., Gillingwater, T.H., Jones, R.A., Loughrey, C.M. and Buisson, N., Smeets, E. and Leonard, H. (2010) Updating the McCarroll, C.S. (2016) Exclusive expression of MeCP2 in the proﬁle of C-terminal MECP2 deletions in Rett syndrome. nervous system distinguishes between brain and peripheral J. Med. Genet., 47, 242–248. 19. Smeets, E., Terhal, P., Casaer, P., Peters, A., Midro, A., Rett syndrome-like phenotypes. Hum. Mol. Genet., 25, 4389–4404. Schollen, E., van Roozendaal, K., Moog, U., Matthijs, G., 4. Skene, P.J., Illingworth, R.S., Webb, S., Kerr, A.R., James, K.D., Herbergs, J. et al. (2005) Rett syndrome in females with CTS hot spot deletions: a disorder proﬁle. Am. J. Med. Genet. A, Turner, D.J., Andrews, R. and Bird, A.P. (2010) Neuronal MeCP2 is expressed at near histone-octamer levels and glob- 132A, 117–120. ally alters the chromatin state. Mol. Cell, 37, 457–468. 20. Guy, J., Gan, J., Selfridge, J., Cobb, S. and Bird, A. (2007) 5. Chen, R.Z., Akbarian, S., Tudor, M. and Jaenisch, R. (2001) Reversal of neurological defects in a mouse model of Rett Deﬁciency of methyl-CpG binding protein-2 in CNS neurons syndrome. Science, 315, 1143–1147. results in a Rett-like phenotype in mice. Nat. Genet., 27, 21. Katz, D.M., Berger-Sweeney, J.E., Eubanks, J.H., Justice, M.J., Neul, J.L., Pozzo-Miller, L., Blue, M.E., Christian, D., Crawley, 327–331. 6. Guy, J., Hendrich, B., Holmes, M., Martin, J.E. and Bird, A. J.N., Giustetto, M. et al. (2012) Preclinical research in Rett syn- (2001) A mouse Mecp2-null mutation causes neurological drome: setting the foundation for translational success. Dis. symptoms that mimic Rett syndrome. Nat. Genet., 27, Model. Mech., 5, 733–745. 322–326. 22. Robinson, L., Guy, J., McKay, L., Brockett, E., Spike, R.C., 7. Christodoulou, J., Grimm, A., Maher, T. and Bennetts, B. Selfridge, J., De Sousa, D., Merusi, C., Riedel, G., Bird, A. et al. (2003) RettBASE: the IRSA MECP2 variation database-a new (2012) Morphological and functional reversal of phenotypes mutation database in evolution. Hum. Mutat., 21, 466–472. in a mouse model of Rett syndrome. Brain, 135, 2699–2710. 8. Lek, M., Karczewski, K.J., Minikel, E.V., Samocha, K.E., Banks, 23. Wegener, E., Brendel, C., Fischer, A., Hulsmann, S., Gartner, J. E., Fennell, T., O’Donnell-Luria, A.H., Ware, J.S., Hill, A.J., and Huppke, P. (2014) Characterization of the MeCP2R168X Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 Human Molecular Genetics, 2018, Vol. 27, No. 14 | 2545 knockin mouse model for Rett syndrome. PLoS One, 9, phenotype of common mutations in Rett syndrome. J. Med. e115444. Genet., 41, 25–30. 24. Brown, K., Selfridge, J., Lagger, S., Connelly, J., De Sousa, D., 38. Baker, S.A., Chen, L., Wilkins, A.D., Yu, P., Lichtarge, O. and Zoghbi, H.Y. (2013) An AT-hook domain in MeCP2 deter- Kerr, A., Webb, S., Guy, J., Merusi, C., Koerner, M.V. et al. (2016) The molecular basis of variable phenotypic severity mines the clinical course of Rett syndrome and related dis- orders. Cell, 152, 984–996. among common missense mutations causing Rett syn- 39. Klose, R.J., Sarraf, S.A., Schmiedeberg, L., McDermott, S.M., drome. Hum. Mol. Genet., 25, 558–570. Stancheva, I. and Bird, A.P. (2005) DNA binding selectivity of 25. Cheval, H., Guy, J., Merusi, C., De Sousa, D., Selfridge, J. and MeCP2 due to a requirement for A/T sequences adjacent to Bird, A. (2012) Postnatal inactivation reveals enhanced re- methyl-CpG. Mol. Cell, 19, 667–678. quirement for MeCP2 at distinct age windows. Hum. Mol. 40. Ghosh, R.P., Nikitina, T., Horowitz-Scherer, R.A., Gierasch, Genet., 21, 3806–3814. L.M., Uversky, V.N., Hite, K., Hansen, J.C. and Woodcock, C.L. 26. Kerr, B., Alvarez-Saavedra, M., Saez, M.A., Saona, A. and (2010) Unique physical properties and interactions of the Young, J.I. (2008) Defective body-weight regulation, motor domains of methylated DNA binding protein 2. Biochemistry, control and abnormal social interactions in Mecp2 hypo- 49, 4395–4410. morphic mice. Hum. Mol. Genet., 17, 1707–1717. 41. Deng, J.V., Wan, Y., Wang, X., Cohen, S., Wetsel, W.C., 27. Samaco, R.C., Fryer, J.D., Ren, J., Fyffe, S., Chao, H.T., Sun, Y., Greenberg, M.E., Kenny, P.J., Calakos, N. and West, A.E. (2014) Greer, J.J., Zoghbi, H.Y. and Neul, J.L. (2008) A partial loss of MeCP2 phosphorylation limits psychostimulant-induced be- function allele of methyl-CpG-binding protein 2 predicts a havioral and neuronal plasticity. J. Neurosci., 34, 4519–4527. human neurodevelopmental syndrome. Hum. Mol. Genet., 17, 42. Li, H., Zhong, X., Chau, K.F., Santistevan, N.J., Guo, W., Kong, 1718–1727. G., Li, X., Kadakia, M., Masliah, J., Chi, J. et al. (2014) Cell 28. Nan, X., Tate, P., Li, E. and Bird, A. (1996) DNA methylation cycle-linked MeCP2 phosphorylation modulates adult neu- speciﬁes chromosomal localization of MeCP2. Mol. Cell. Biol., rogenesis involving the Notch signalling pathway. Nat. 16, 414–421. Commun., 5, 5601. 29. Nan, X., Ng, H.H., Johnson, C.A., Laherty, C.D., Turner, B.M., 43. Li, H., Zhong, X., Chau, K.F., Williams, E.C. and Chang, Q. Eisenman, R.N. and Bird, A. (1998) Transcriptional repression (2011) Loss of activity-induced phosphorylation of MeCP2 by the methyl-CpG-binding protein MeCP2 involves a his- enhances synaptogenesis, LTP and spatial memory. Nat. tone deacetylase complex. Nature, 393, 386–389. Neurosci., 14, 1001–1008. 30. Lyst, M.J., Ekiert, R., Ebert, D.H., Merusi, C., Nowak, J., 44. Zhou, Z., Hong, E.J., Cohen, S., Zhao, W.N., Ho, H.Y., Schmidt, Selfridge, J., Guy, J., Kastan, N.R., Robinson, N.D., de Lima L., Chen, W.G., Lin, Y., Savner, E., Grifﬁth, E.C. et al. (2006) Alves, F. et al. (2013) Rett syndrome mutations abolish the in- Brain-speciﬁc phosphorylation of MeCP2 regulates teraction of MeCP2 with the NCoR/SMRT co-repressor. Nat. activity-dependent Bdnf transcription, dendritic growth, Neurosci., 16, 898–902. and spine maturation. Neuron, 52, 255–269. 31. Gofﬁn, D., Allen, M., Zhang, L., Amorim, M., Wang, I.T., 45. Cheng, T.L., Wang, Z., Liao, Q., Zhu, Y., Zhou, W.H., Xu, W. Reyes, A.R., Mercado-Berton, A., Ong, C., Cohen, S., Hu, L. and Qiu, Z. (2014) MeCP2 suppresses nuclear microRNA proc- et al. (2012) Rett syndrome mutation MeCP2 T158A disrupts essing and dendritic growth by regulating the DNA binding, protein stability and ERP responses. Nat. DGCR8/Drosha complex. Dev. Cell, 28, 547–560. Neurosci., 15, 274–283. 46. Lamonica, J.M., Kwon, D.Y., Gofﬁn, D., Fenik, P., Johnson, 32. Pitcher, M.R., Herrera, J.A., Bufﬁngton, S.A., Kochukov, M.Y., B.S., Cui, Y., Guo, H., Veasey, S. and Zhou, Z. (2017) Elevating Merritt, J.K., Fisher, A.R., Schanen, N.C., Costa-Mattioli, M. expression of MeCP2 T158M rescues DNA binding and Rett and Neul, J.L. (2015) Rett syndrome like phenotypes in the syndrome-like phenotypes. J. Clin. Invest., 127, 1889–1904. R255X Mecp2 mutant mouse are rescued by MECP2 trans- 47. Bain, G., Kitchens, D., Yao, M., Huettner, J.E. and Gottlieb, D.I. gene. Hum. Mol. Genet., 24, 2662–2672. (1995) Embryonic stem cells express neuronal properties 33. Cuddapah, V.A., Pillai, R.B., Shekar, K.V., Lane, J.B., Motil, K.J., in vitro. Dev. Biol., 168, 342–357. Skinner, S.A., Tarquinio, D.C., Glaze, D.G., McGwin, G., 48. Li, M., Pevny, L., Lovell-Badge, R. and Smith, A. (1998) Kaufmann, W.E. et al. (2014) Methyl-CpG-binding protein 2 Generation of puriﬁed neural precursors from embryonic (MECP2) mutation type is associated with disease severity in stem cells by lineage selection. Curr. Biol., 8, 971–974. Rett syndrome. J. Med. Genet., 51, 152–158. 49. Scholz, D., Poltl, D., Genewsky, A., Weng, M., Waldmann, T., 34. Lotharius, J., Barg, S., Wiekop, P., Lundberg, C., Raymon, H.K. Schildknecht, S. and Leist, M. (2011) Rapid, complete and and Brundin, P. (2002) Effect of mutant alpha-synuclein on large-scale generation of post-mitotic neurons from the hu- dopamine homeostasis in a new human mesencephalic cell man LUHMES cell line. J. Neurochem., 119, 957–971. line. J. Biol. Chem., 277, 38884–38894. 50. Schwenk, F., Baron, U. and Rajewsky, K. (1995) A 35. Shah, R.R., Cholewa-Waclaw, J., Davies, F.C.J., Paton, K.M., cre-transgenic mouse strain for the ubiquitous deletion of Chaligne, R., Heard, E., Abbott, C.M. and Bird, A.P. (2016) loxP-ﬂanked gene segments including deletion in germ cells. Efﬁcient and versatile CRISPR engineering of human neu- Nucleic Acids Res., 23, 5080–5081. rons in culture to model neurological disorders. Wellcome 51. Aida, T., Chiyo, K., Usami, T., Ishikubo, H., Imahashi, R., Open Res., 1, 13. Wada, Y., Tanaka, K.F., Sakuma, T., Yamamoto, T. and 36. Bebbington, A., Anderson, A., Ravine, D., Fyfe, S., Pineda, M., Tanaka, K. (2015) Cloning-free CRISPR/Cas system facilitates de Klerk, N., Ben-Zeev, B., Yatawara, N., Percy, A., functional cassette knock-in in mice. Genome Biol., 16, 87. Kaufmann, W.E. et al. (2008) Investigating 52. Schaefer, M.R., Williams, M., Kulpa, D.A., Blakely, P.K., genotype-phenotype relationships in Rett syndrome using Yaffee, A.Q. and Collins, K.L. (2008) A novel trafﬁcking signal an international data set. Neurology, 70, 868–875. within the HLA-C cytoplasmic tail allows regulated expres- 37. Colvin, L., Leonard, H., de Klerk, N., Davis, M., Weaving, L., sion upon differentiation of macrophages. J. Immunol., 180, Williamson, S. and Christodoulou, J. (2004) Reﬁning the 7804–7817. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Human Molecular Genetics Oxford University Press http://www.deepdyve.com/lp/oxford-university-press/a-mutation-led-search-for-novel-functional-domains-in-mecp2-D9KkulrC2V

Loading next page...

Publisher: Oxford University Press
Copyright: Copyright © 2022 Oxford University Press
ISSN: 0964-6906
eISSN: 1460-2083
DOI: 10.1093/hmg/ddy159
Publisher site: See Article on Publisher Site

Abstract

Most missense mutations causing Rett syndrome (RTT) affect domains of MeCP2 that have been shown to either bind methylated DNA or interact with a transcriptional co-repressor complex. Several mutations, however, including the C-termi- nal truncations that account for 10% of cases, fall outside these characterized domains. We studied the molecular conse- quences of four of these ‘non-canonical’ mutations in cultured neurons and mice to see if they reveal additional essential domains without affecting known properties of MeCP2. The results show that the mutations partially or strongly deplete the protein and also in some cases interfere with co-repressor recruitment. These mutations therefore impact the activity of known functional domains and do not invoke new molecular causes of RTT. The ﬁnding that a stable C-terminal truncation does not compromise MeCP2 function raises the possibility that small molecules which stabilize these mutant proteins may be of therapeutic value. Introduction polymorphisms are broadly distributed across the remainder of Mutations in the MECP2 gene cause the profound neurological the protein (Fig. 1). The implication that RTT mutational disorder Rett syndrome (1) (RTT, OMIM #312750). The gene prod- clusters mark critical domains is supported by biochemical uct, MeCP2 protein, is expressed in most cell types, but is espe- studies, which show that mutations within the Methyl-CpG cially abundant in neurons (2–4). Accordingly, loss of MeCP2 binding domain (MBD) disrupt DNA binding, whereas mutations function has the most profound effect in the nervous system, within the NCoR interaction domain (NID) prevent binding to with minimal phenotypic consequences in other tissues (3,5,6). the NCoR/SMRT co-repressor complexes (9–11). High-resolution Because of its association with intellectual disability in RTT and X-ray crystallographic structures of both these interactions other disorders, the MECP2 gene is frequently screened for support the view that MeCP2 bridges methylated sites on DNA mutations in clinical cases of developmental delay. This has de- with a histone deacetylase-containing co-repressor (12,13). ﬁned many amino acid changes as RTT-causing or as relatively Further evidence for the importance of these two domains benign or neutral variants that do not cause RTT (7,8). The comes from a recent study in which mice carried a pared down MeCP2 protein sequence is 95% identical between rodents and MeCP2 ‘minigene’, DNIC, consisting of only the MBD, a nuclear humans, suggesting stringent purifying selection throughout. localization signal and the NID (14). These mice survived for Nevertheless, RTT missense mutations are non-randomly dis- over a year with only mild symptoms. A simple interpretation is tributed in two prominent clusters, whereas apparently neutral that this interaction restrains transcription; a view supported Received: March 7, 2018. Revised: April 24, 2018. Accepted: April 26, 2018 V C The Author(s) 2018. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. 2531 Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 2532 | Human Molecular Genetics, 2018, Vol. 27, No. 14 ExAC variants 1 486 MeCP2 MBD MBD P225R P322L * * RTT missense CTDs RTT truncations Figure 1. Population variants and RTT mutations in the coding sequence of MeCP2. A schematic diagram of human MeCP2 protein shows the methyl-binding domain (MBD) and NCoR/SMRT interaction domain (NID). Single amino acid changes found in males in the ExAC exome sequence database are indicated above the protein and RTT-causing mutations are shown below. Only male ExAC changes are shown to exclude the possibility that pathogenic mutations in females could be masked by skewed X-inactivation. All RTT mutations have been shown to be absent in the parents in at least one case. Red box—range of deletion start points for C-terminal dele- tion (CTD) RTT frameshift mutations. The ‘non-canonical’ RTT mutations that are the subject of this study are marked with asterisks. by recent transcriptomic analyses of MeCP2-depleted mouse MBD and the NID. According to our rationale, new functional brain (15–17). domains would be expected to cause RTT without compromis- A signiﬁcant limitation of this model for MeCP2 function is ing either binding to DNA or NCoR recruitment. that several recurrent mutations, whose causal role in RTT is We initially selected the P225R and P322L mutations, which conﬁrmed by their absence in both parents, lie outside the MBD occur in a relatively small number of RTT cases (Table 1). Parental analysis has shown that these missense mutations and NID (Fig. 1) and are absent in DNIC mice (14). Most promi- nently, about 10% of all RTT cases involve a heterogeneous set arise de novo in the affected individuals (7), strongly suggesting that they are causative. We modeled these mutations by gene of small deletions in exon 4 of MECP2. The resulting frameshifts introduce premature translational termination often preceded targeting in mouse embryonic stem (ES) cells (Fig. 2a; Supplementary Material, Fig. S1a) to produce mouse alleles that by stretches of novel amino acid sequence of variable length differed from WT only at the patient mutation site and a single (18,19). As these C-terminal deletions (CTDs) are far down- loxP site remaining in intron 2 after deletion of the selection stream of the MBD and NID, they hint that there are other, as cassette (Fig. 2a; Supplementary Material, Fig. S1b–e). To deter- yet undetected, functional domains within MeCP2. The same ar- mine whether P225R and P322L cause RTT-like phenotypes in gument applies to two other recurrent RTT mutations, P225R mice, we injected mutant ES cells into blastocysts to create chi- and P322L, which lie outside the minimal regions required for P225R meras and bred them to establish Mecp2 (P225R) and DNA and NCoR binding. Here we use biochemical, cell biological P322L Mecp2 (P322L) mouse lines (Supplementary Material, Fig. and genetic methods to show that these mutations recapitulate S2a and b). P322L hemizygous males appeared normal at wean- RTT-like phenotypes in mice through either or both of two ing, but from 5 weeks of age they developed RTT-like symp- mechanisms: 1) by destabilizing MeCP2, so that only a fraction toms (20) that progressed rapidly (Fig. 2b), surviving to a median of the wild-type (WT) level remains; 2) by interfering with the age of 9 weeks (Fig. 2c). P322L males were also underweight ability of MeCP2 to recruit the NCoR complex to chromatin and compared to littermate controls (Supplementary Material, Fig. repress transcription. Importantly, we ﬁnd that one C-terminal S3a). These mice were subjected to a program of behavioral truncation is stable and functionally WT in a mouse sequence tests starting at 6 weeks old when the mutants were only mildly context, but unstable and RTT-causing when the protein is symptomatic. They showed signiﬁcantly decreased anxiety-like ‘humanized’. This demonstrates that the truncation itself behaviors in the elevated plus maze and open ﬁeld test only has a measurable functional consequence if it leads to de- (Supplementary Material, Fig. S3b and c) although their move- stabilization. From this and other data we conclude that these ment was not signiﬁcantly impaired (data not shown and ‘non-canonical’ RTT mutations do not identify previously unap- Supplementary Material, Fig. S3d). The wire suspension test and preciated functional domains. Moreover, our results highlight accelerating rotarod (Supplementary Material, Fig. S3e and f) ex- the crucial role played by the MeCP2 protein sequence in ensur- posed severe deﬁcits in motor co-ordination and learning. Thus ing stability. P322L males exhibit a severe phenotype mimicking that of Mecp2-null mice. Results As with P322L mice, P225R males developed Rett-like symp- toms at 6 weeks, but progression of symptoms was more grad- P225R and P322L mutations cause Rett-like phenotypes ual (Fig. 2d). Median survival was 50 weeks (Fig. 2e). P225R in mice males were underweight compared to WT littermates To test the hypothesis that the MBD and NID are key to MeCP2 (Supplementary Material, Fig. S4a). Despite their extended sur- function, we ﬁrst identiﬁed RTT mutations that apparently vival time (the longest of any Rett mutation modeled in mice so leave these domains intact. Next, we generated mouse models far), P225R mice showed a similar range of deﬁcits to P322L in to check that these recapitulated RTT-like phenotypes. Finally, behavioral tests. P225R males at 11 weeks of age appeared less we used cellular and molecular assays to ask whether these anxious than controls in the elevated plus maze and open ﬁeld mutations affect the functions that have been attributed to the test (Supplementary Material, Fig. S4b and c). They traveled a NID NID Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 Human Molecular Genetics, 2018, Vol. 27, No. 14 | 2533 (a) 12 3 4 Mecp2 genomic locus 30kb plasmid subclone targeting Neo Stop vector loxP loxP RTT mutation targeted genomic locus loxP (b) (c) P322L P322L WT WT null null 0 0 4 8 12 16 20 24 28 0 4 8 1216202428 Age (weeks) Age (weeks) (d) (e) 10 100 P225R WT 8 flox 80 null 6 60 P225R WT 4 40 null 2 20 0 0 4 8 12 16 20 24 28 32 36 40 44 48 52 0 4 8 1216202428323640444852 Age (weeks) Age (weeks) Figure 2. P322L and P225R mutant mice show RTT-like phenotypes and a reduced lifespan. (a) The mouse Mecp2 genomic locus, the targeting vector used to make P225R and P322L knock-in alleles in mouse ES cells, and the mutated genomic locus after Cre-mediated removal of the selection cassette. The region where the RTT mutations were introduced is marked with an asterisk (*). (b) P322L hemizygous male mice show a rapid onset of RTT-like phenotypes. Mean aggregate phenotypic score6SD plotted for each genotype, P322L n¼12 at 4 weeks, WT littermates n¼15. Representative Mecp2-null data are shown for comparison (n¼12 at 4 weeks). (c) Kaplan–Meier survival plot for P322L hemizygous males and WT littermates shown in (b). (d) P225R hemizygous male mice also display RTT-like phenotypes. Mean ag- gregate phenotypic score6SD plotted for each genotype, P225R n¼19 at 4 weeks, WT littermates n¼18. Representative Mecp2-null and Mecp2-ﬂox data are shown for comparison (ﬂox n¼9). (e) Kaplan–Meier survival plot for P225R hemizygous males and WT littermates shown in (d). Table 1. MECP2 mutations in this study. Amino acid and nucleotide locations are numbered according to the human e2 isoform. RettBASE fre- quency is expressed as a percentage of all Rett syndrome cases with an MECP2 mutation Mutation name Amino acid change Nucleotide change Mutation type RettBASE frequency MeCP2 domain R111G R111G c.331 A>G RTT missense 0.03 MBD P225A P225A c.673 C>G ExAC (male) n/a ID P225R P225R c.674 C>G RTT missense 0.55 ID R306C R306C c.916 C>T RTT missense 7.3 NID P322L P322L c.965 C>T RTT missense 0.26 CTD CTD1 L386HfsX5 c.1157–1197 D41 RTT del/frameshift 1.32 CTD CTD2 P389X c.1164–1207 D44 RTT del/frameshift 1.29 CTD MBD, methyl-CpG binding domain, ID, intervening domain (between the MBD and NID), NID, NCoR interacting domain and CTD, C-terminal domain. Aggregate Score Aggregate Score Percent survival Percent survival Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 2534 | Human Molecular Genetics, 2018, Vol. 27, No. 14 shorter distance in the open ﬁeld (Supplementary Material, Fig. heterochromatic foci containing high concentrations of methyl- S4d) and also demonstrated reduced motor co-ordination in the ated DNA (28). A mutation that disrupts the MBD, R111G, abol- wire suspension test (Supplementary Material, Fig. S4e), though ished binding of MeCP2 to methylated DNA and co-localization was lost. Both P225R and P322L (in addition to controls P225A mutants did not perform less well than controls on the acceler- ating rotarod (Supplementary Material, Fig. S4f). We conclude and R306C) retained the ability to bind to heterochromatic foci, indicating that MBD function is not affected in these mutants that both P322L and P225R male mice show phenotypes (Fig. 4a). that overlap with those of other mouse models carrying Mecp2 MeCP2 interacts with the co-repressor complex NCoR/SMRT mutations found in RTT (21–24). via subunits TBL1X and TBL1XR1. This interaction is disrupted by mutations in the NCoR/SMRT interaction domain, including P225R and P322L mutations differentially affect MeCP2 the common Rett mutation R306C. EGFP-tagged MeCP2 was levels expressed in HEK 293 cells and immunoprecipitated from 150 mM salt extracts using GFP-Trap beads (Supplementary To investigate the molecular basis of these RTT-like pheno- Material, Fig. S6a). NCoR/SMRT components NCoR1, TBL1X and types, we ﬁrst measured MeCP2 abundance. Strikingly, the HDAC3 co-immunoprecipitated with WT MeCP2 but not with amount of P322L protein in whole mouse brain was severely re- the R306C mutant or EGFP alone. Both P225R and P322L mutants duced (3% of WT littermates; Fig. 3a and b). Thus, despite WT P322L were able to immunoprecipitate the same NCoR/SMRT compo- amounts of Mecp2 mRNA (Supplementary Material, Fig. S5a nents as the WT protein (Fig. 4b), showing that these mutations and b), the level of P322L protein in brain is comparable to that Stop/y do not impair the interaction of MeCP2 with the co-repressor in Mecp2 mice, which also show a severe phenotype (22). In complex. The co-repressor mSin3a, which interacts relatively P225R mouse brain, MeCP2 protein was at 22% of the WT level weakly with MeCP2 (29), was immunoprecipitated by all mutant (Fig. 3c and d), which is signiﬁcantly more than in P322L MeCP2 proteins, including R306C (Supplementary Material, Fig. P225R brain. The amount of Mecp2 mRNA was not affected S6b). (Supplementary Material, Fig. S5b). Although P225R and P322L can still interact with NCoR and Reduced MeCP2 protein abundance was also observed in cul- methylated DNA separately, we wished to know if they could tured neurons derived from the mutant mouse ES cells (Fig. 3e– recruit the NCoR subunit TBL1X to heterochromatin, which g). Both P322L and P225R ES cells differentiated equally well into requires simultaneous binding to both macromolecules. To test neurons, as shown by NeuN expression (Fig. 3f) and displayed this, constructs expressing both EGFP-tagged MeCP2 and equal amounts of Mecp2 mRNA (Supplementary Material, Fig. mCherry-tagged TBL1X were transfected into mouse NIH-3T3 S5c). However, protein levels in both mutants were signiﬁcantly ﬁbroblasts (Supplementary Material, Fig. S7a–e). It has been lower than WT and WT þ lox controls (Fig. 3f and g). shown previously (30) that TBL1X-mCherry, which lacks a nu- Speciﬁcally, P225R showed a moderate reduction to around 60% clear localization signal, is normally excluded from the nucleus whereas P322L retained only 7% of the WT MeCP2 level. (Supplementary Material, Fig. S7b), but in the presence of WT EGFP-MeCP2 it is recruited to the methylation-rich 4’, 6-diami- dino-2-phenylindole (DAPI) spots in 80% of doubly transfected P225R and P322L impair NCoR recruitment to chromatin cells (Fig. 4c and d). As expected, the NID mutant R306C itself The drastic reduction in P322L MeCP2 explains the severe phe- localizes to the DAPI spots, but is unable to recruit TBL1X notype, but the P225R mutant is only partially depleted. (Fig. 4c). When P225R and P322L RTT mutants were tested in Notably, the level of P225R MeCP2 in whole brain is similar to this assay both showed impaired recruitment of TBL1X- that of the hypomorphic ﬂoxed Mecp2 allele (6,25–27)(Fig. 3c), mCherry to DAPI spots in comparison with WT MeCP2. On the which has a normal lifespan and much milder phenotypic pro- other hand, the population variant P225A performed as well as gression than P225R (25–27) [data from (25) shown for compari- WT MeCP2 in this assay (Fig. 4d; Supplementary Material, Fig. son in Fig. 2d, yellow symbols]. The difference between S7c). Thus the ability of both P225R and P322L to recruit NCoR/ phenotypic outcomes may indicate that these two mouse lines SMRT to methylated sites in the genome is impaired. Though simply lie on either side of a threshold for functionally sufﬁcient performed in non-neuronal cells, this test appears to measure a MeCP2 protein levels. Alternatively, the P225R mutation may af- property of MeCP2 that is functionally relevant, as Rett muta- fect MeCP2 function as well as abundance. To test the possibil- tions register as defective. Strikingly, substitution of R at posi- ity that the residual mutant protein is functionally impaired, we tion 225 causes RTT and prevents recruitment, whereas looked for other defects that might account for the RTT-like substitution of A at the same site does not cause RTT and phenotype. While many of the effects of MeCP2 mutations only recruits as efﬁciently as WT. become apparent when studied in the context of the whole or- Failure to efﬁciently recruit the cognate co-repressor to ganism, some properties, such as recruitment of the NCoR/ methylated sites in the genome suggested that P225R and P322L SMRT co-repressor complex to chromatin, can be tested using a might also be functionally defective in DNA methylation- simple transfection assay. P225R and P322L point mutations dependent transcriptional repression. To test this, untagged were introduced into a WT mouse Mecp2 cDNA fused to the C- MeCP2 was expressed in mouse tail ﬁbroblasts that were com- terminus of enhanced green ﬂuorescent protein (EGFP). promised in their ability to repress methylated reporter genes Interestingly, the Exome Aggregation Consortium (ExAC) data- due to deletion of the genes for Mecp2 and Mbd2 as previously base identiﬁes a hemizygous missense population variant also described (6). The co-transfected reporter construct encoding at P225 (P225A) that presumably does not cause RTT. We in- ﬁreﬂy luciferase was either unmethylated or methylated at ev- cluded this variant in our analysis and used Rett missense ery CpG (Fig. 4e) and repression was expressed as the ratio of lu- mutations R306C and R111G as negative controls for NCoR and ciferase activities expressed by unmethylated versus chromatin binding, respectively (Table 1). When WT EGFP- methylated constructs. Equivalent expression levels of WT, MeCP2 was expressed in mouse NIH-3T3 cells, green ﬂuores- R306C and P225R MeCP2 constructs were conﬁrmed by western cence co-localized with ‘DAPI bright spots’, which are blotting (Supplementary Material, Fig. S8a and b). Levels of Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 Human Molecular Genetics, 2018, Vol. 27, No. 14 | 2535 (a)(b) WT P322L null Stop MeCP2 H3 ** WT P322L (c) (d) WT P225R null flox MeCP2 H3 WT P225R (e) ES cells EBs -RA EBs +RA NPs neurons day 0 4 8 15 -LIF +RA plating harvest (f) (g) P225R P322L WT+lox WT 160 ns AABB AB pools MeCP2 100 NeuN H3 *** Figure 3. P322L and P225R mutations result in decreased MeCP2 levels in mouse brain and in vitro differentiated neurons. (a) Western blot of whole brain lysates from 6-week-old P322L and WT littermate males. Mecp2-null and Mecp2-stop samples are shown for comparison. (b) Quantiﬁcation of (a): the level of MeCP2 (MeCP2 signal/ H3 signal) as a percentage of mean WT is shown as mean (black line) and individual biological replicates (n¼3 animals). P¼0.0010 (**), two-tailed t-test with Welch’s cor- rection for unequal variances. (c) Western blot of whole brain lysates from 9-week-old P225R and WT littermate males. Mecp2-null and Mecp2-ﬂox samples are shown for comparison. (d) Quantiﬁcation of data shown in (c): the level of MeCP2 (MeCP2 signal/H3 signal) as a percentage of mean WT is shown as mean (black line) and indi- vidual biological replicates (n¼3 animals). P¼0.0128 (*), two-tailed unpaired t-test with Welch’s correction for unequal variances. (e) Neuronal differentiation protocol used to produce MeCP2-mutant neurons from mouse ES cells. EB: embryoid body, RA: retinoic acid, NP: neuronal precursor. (f) Representative western blot showing lev- els of MeCP2 protein present in neurons on day 15 of the differentiation scheme. Lanes A and B for each genotype are derived from independently targeted ES cell clones and WT pools are independent differentiations of the parental ES cell line. Histone H3 was used as a loading control for number of cells and NeuN as a control for equal neuronal differentiation. (g) Quantiﬁcation of MeCP2 levels in in vitro differentiated neurons. Two independent clones were differentiated two or three times for each genotype. MeCP2 level (MeCP2 signal/H3 signal) is normalized to the mean WTþloxP value. Data shown as mean (black line) and individual values. n¼5 (WT þloxP, P225R, P322L), n¼3 (WT pool) independent differentiations. Comparison to WTþloxP (two-tailed unpaired t-test with Welch’s correction for unequal variances): P225R P¼0.0416 (*), P322L P¼0.0023 (***) and WT (pool) P¼0.4868 (ns). P322L, however, were consistently low (30% of WT MeCP2; that the residual repression seen with R306C is due to recruit- data not shown), suggesting that lack of an EGFP tag rendered ment of another co-repressor, such as mSin3a, or to direct inter- the protein susceptible to degradation, as seen in neurons. As ference of dense methyl-CpG bound protein with the the amount of MeCP2 directly correlates with the strength of re- transcriptional machinery. pression in this assay, we excluded the P322L mutant from the transcriptional repression experiment. WT MeCP2 repressed the C-terminal deletions do not affect MeCP2 function or methylated luciferase efﬁciently, whereas repression by P225R recruitment was reduced by about half. The ability of R306C to repress was about 25% that of WT MeCP2 in this assay (Fig. 4f). These ﬁnd- Deletion-frameshift mutations in the C-terminal domain of ings correspond well with the behavior of the mutants in the MECP2, generally between nucleotides c.1100 and c.1200 (human TBL1X recruitment assay, where R306C was more severely af- e2 isoform numbering), are responsible for about 10% of all RTT fected than P225R (and P322L), supporting the view that failure cases. We chose to model two of the most common conﬁrmed to recruit the NCoR co-repressor underlies DNA methylation- Rett-causing CTDs in mice: c.1157–1197 D41 (CTD1) and c.1164– dependent transcriptional repression by MeCP2. It is possible 1207 D44 (CTD2; see Table 1). Native mouse MeCP2 contains a two WT+loxP P225R P322L WT pool MeCP2 (% WT+loxP) % WT mean % WT mean Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 2536 | Human Molecular Genetics, 2018, Vol. 27, No. 14 (a) WT R111G R306C P225A P225R P322L DAPI EGFP merge 10μm (b) EGFP WT R306C P225R P322L in IP in IP in IP in IP in IP NCoR1 300kD TBL1X 50kD 50kD HDAC3 (c) (d) EGFP mCherry DAPI merge ns WT ** R306C ** (e)(f) SV E 80 Luciferase ORF pA SV **** CpG **** **** Figure 4. P225R and P322L show impaired NCoR/SMRT recruitment to methylated DNA. (a) WT and mutant EGFP-MeCP2 fusions expressed in mouse NIH-3T3 cells. (b) Immunoprecipitation of EGFP-MeCP2 from transfected HEK 293 cells with GFP-Trap beads. NCoR/SMRT components NCoR1, TBL1X and HDAC3 detected in western blots of input and immunoprecipitated samples. (c) TBL1X recruitment assay. Co-transfection of EGFP-MeCP2 and TBL1X-mCherry expression constructs into NIH-3T3 cells. WT EGFP-MeCP2, but not R306C, recruits TBL1X-mCherry to heterochromatic foci. Scale bar 10 lm. (d) Quantiﬁcation of TBL1X recruitment efﬁciency of MeCP2 mutants. Percent of doubly transfected nuclei with TBL1X-mCherry/EGFP-MeCP2 spots from each independent transfection is shown as mean (black line) and individ- ual values. n¼7 (WT), 4 (R306C, P322L), 3 (P225A, P225R) independent transfections. MeCP2 mutants were compared to WT using a two-tailed Mann–Whitney test. R306C P¼0.0061 (**), P225A P¼0.1167 (ns), P225R P¼0.0167 (*), P322L P¼0.0061 (**). (e) Diagram of ﬁreﬂy luciferase expression construct pGL2-control consisting of an SV40 promoter, Luc coding sequence and SV40 polyadenylation signal and enhancer. The positions of CpG dinucleotides methylated by SssI.methylase are shown as black /y / lines. (f) Methylation-dependent transcriptional repression assay. Immortalized Mecp2 , Mbd2 mouse tail ﬁbroblasts transfected with SssI-methylated or unmethy- lated pGL2-control plasmid and an MeCP2 expression construct. Methylation-dependent repression expressed as the ratio of unmethylated/methylated (U/M) lucifer- ase activity and normalized to WT for each transfection. For each genotype the mean (black line) and individual values for four (R306C) or six (no MeCP2, WT, P225R) independent transfection experiments are shown. Each genotype compared to the WT value (100%) using a two-tailed one-sample t-test. No MeCP2 P<0.0001 (****), R306C P<0.0001 (****), P225R P¼0.0002 (****). WT R306C P225A P225R P322L WT R306C P225R U/M luc activity (%WT) % cells TBL1X recruited Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 Human Molecular Genetics, 2018, Vol. 27, No. 14 | 2537 amino acid deletion relative to the human protein (Fig. 5a)that mRNA levels in whole brain were not signiﬁcantly different coincides with the deletion hotspot in Rett patients. We modeled from WT littermate controls (Fig. 6a and b; Supplementary these Rett mutations in the mouse genome by restoring the two Material, Fig. S10e). Strikingly, CTD2 male hemizygous mice and human-speciﬁc amino acids and by adding the patient missense WT littermate controls were phenotypically indistinguishable tail (CTD1: HQPPX, CTD2: X) (Fig. 5a). We ﬁrst tested CTD1 and over the course of 1 year (Fig. 6e and f). Survival was 100% and CTD2 for methylated DNA-binding and NCoR/SMRT binding us- CTD2 mice behaved as WT in the elevated plus maze (Fig. 6g), ing transfection assays. Both were able to bind to methylated het- open ﬁeld test (Fig. 6h; Supplementary Material, Fig. S10f) and erochromatic foci in transfected NIH-3T3 cells (Fig. 5b) and to co- accelerating rotarod (Fig. 6i). In the wire suspension test CTD2 immunoprecipitate NCoR/SMRT components NCoR1, TBL1X and mice performed signiﬁcantly better than WT littermates HDAC3 (Fig. 5c) and mSin3a (Supplementary Material, Fig. S6b). (Supplementary Material, Fig. S10g). We conclude that the CTD2 Moreover, unlike P225R and P322L, both CTD1 and CTD2 were C-terminal truncation of MeCP2 has no overt phenotypic conse- able to recruit TBL1X-mCherry to methylated heterochromatic quences in this mouse model and therefore does not identify a foci as efﬁciently as WT MeCP2 (Fig. 5d). Thus, neither of the trun- previously unknown functional domain. cated proteins showed any deﬁciencies in MBD or NID function in these assays. ‘Humanization’ of the CTD2 allele reduces protein levels in neurons CTD1 and CTD2 mutant phenotypes reﬂect MeCP2 All Rett mutations previously studied in mice have replicated protein level the disease seen in patients, including the relative severity of The absence of overt defects in these assays raised the possibil- symptoms (5,6,23,24,31–33). Why should CTD2 be exceptional? ity that CTD1 and CTD2 disrupt other functional domains. To A potential explanation was that subtle differences in amino explore this possibility, we used CTD1 and CTD2 targeting vec- acid sequence between human and mouse MeCP2 proteins tors that retained all sequences downstream of the deletions to were involved. Despite 95% identity of MeCP2 protein between CTD1/y CTD2/y create Mecp2 and Mecp2 ES cell lines (Supplementary the two species, several amino acid differences persisted near Material, Fig. S9a and b). After Cre-mediated deletion of the the truncated region (Fig. 7a). To test whether these differences ﬂoxed Neo-Stop selection cassette, the mutated Mecp2 alleles altered stability of the mutant, we created a ‘humanized’ CTD2 differed from WT only at the mutation site and a loxP site up- allele (CTD2hu) in mouse ES cells by making three single nucle- stream of exon 3. When these lines were differentiated into otide alterations: T376P, M380V and S387P (Fig. 7a). These neurons in vitro, both genotypes produced truncated MeCP2 pro- changes ensure that the amino acid sequence of the C-terminus tein as expected, but surprisingly the amount of protein differed of MeCP2 in patient and mouse alleles is identical from E298, greatly. CTD1 retained only 3% the WT level, whereas the which is N-terminal to the NID, to the stop codon. CTD2hu ES amount of CTD2 MeCP2 was not signiﬁcantly different from WT cell clones were produced (Supplementary Material, Fig. S11a controls (Fig. 5e and f). At the transcript level, CTD1 had only and b) and differentiated into neurons for comparison with 34% WT mRNA whereas CTD2, like P225R and P322L, expressed CTD1, CTD2 and WT þ lox controls. Western blots showed a WT-levels of mRNA (Supplementary Material, Fig. S9c). dramatic destabilization of CTD2hu, which was now indistin- The extremely low level of MeCP2 protein in CTD1 neurons guishable from CTD1 (7.4 and 5.0% WT þ lox, respectively, predicted that mice would show a severe phenotype, similar to Fig. 7b and c). The original CTD2 MeCP2 protein remained essen- P322L and Mecp2-null mice. To test this, we produced CTD1 mice tially identical to WT þ lox controls (93%). CTD2hu neurons also using CRISPR/Cas9 cutting and repair with an oligonucleotide contained less mRNA than WT þ lox and CTD2, again resem- template in mouse zygotes. After conﬁrming the presence of the bling CTD1 (Supplementary Material, Fig. S11c). Thus, altering mutation by sequencing (data not shown) and southern blot the sequence context of the non-pathogenic CTD2 mouse allele (Supplementary Material, Fig. S10a and b), analysis of CTD1 to make it resemble the patient mutation more closely resulted founder males showed that the amount of MeCP2 in whole brain in a drastic reduction in the amount of MeCP2 in neurons. was reduced to about 10% of that in WT controls (Fig. 6a and b) To further conﬁrm our hypothesis that CTD2 MeCP2 is unsta- and the mRNA to 45% of WT (Supplementary Material, Fig. S10e). CTD2 ble in human neurons we created a knock-in MECP2 allele in A cohort of CTD1 male mice was scored for Rett-like phenotypes, Lund human mesencephalic cells (LUHMES) cells, a human neuro- which progressed rapidly (Fig. 6c), giving a median survival of 20 nal progenitor cell line derived from human embryonic ventral weeks (Fig. 6d). The weights of CTD1 mice were not signiﬁcantly mesencephalic tissue which can be differentiated in vitro to pro- different to those of WT littermates (Supplementary Material, Fig. duce a uniform population of mature dopaminergic neurons S10c). This cohort was still on a mixed CBA/C57BL6 genetic back- CTD2 (34,35). Knock-in clones containing MECP2 alleles on the active ground whereas small size is usually seen in Mecp2-mutant males X chromosome, and thus expressing MeCP2 CTD2 protein, were on a more inbred C57BL6 background. produced using CRISPR/Cas9 cutting and an oligodeoxynucleotide The instability of CTD1 made it impossible to deduce repair template as previously described (35)(Supplementary whether the C-terminal truncation mutation gave rise to a novel Material,Fig.S11dand e). After differentiating for 9 days to mature functional defect in MeCP2 as at 10% of WT littermate controls neurons, the level of CTD2 MeCP2 protein was severely reduced the level of protein was similar to the level of full-length protein Stop/y compared to the WT parental cells and unmodiﬁed WT clones in Mecp2 mice, which resemble Mecp2-null mice (22). On the (Fig. 7d and e). MeCP2 protein levels in CTD2 clones ranged from 7 other hand, the fully stable CTD2 mutant protein offered the to 20% of the parental WT cells. These ﬁndings in mice, mouse ES chance to test the importance of this C-terminal region. CTD2 mice were therefore produced using standard injection of ES cell-derived neurons and human neurons strongly suggest that cells into mouse blastocysts and veriﬁed by southern blots RTT resulting from CTD mutations is due to severe MeCP2 deﬁ- (Supplementary Material, Fig. S10d) and sequencing (data not ciency rather than the loss of essential domains in the C-terminal shown). As in cultured neurons, CTD2 MeCP2 protein and region. Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 2538 | Human Molecular Genetics, 2018, Vol. 27, No. 14 (a) (b) (d) (c) (e) (f) Figure 5. C-terminal truncation mutants CTD1 and CTD2 show normal NCoR/SMRT recruitment but differ in neuronal protein levels. (a) Comparison of the human and mouse MeCP2 CTD regions for wild-type (WT) and the patient mutations CTD1 and CTD2. Amino acid changes and absent amino acids in the mouse protein are shown in red, amino acids added to mouse CTD1 and CTD2 proteins in blue and missense/nonsense changes arising due to deletion/frameshift in bold type. (b) WT, CTD1 and CTD2 EGFP-MeCP2 fusions expressed in mouse NIH-3T3 cells. (c) Immunoprecipitation of EGFP-MeCP2 from transfected HEK 293 cells with GFP-Trap beads. NCoR/ SMRT components NCoR1, TBL1X and HDAC3 detected in western blots of input and immunoprecipitated samples. (d) Quantiﬁcation of TBL1X recruitment efﬁciency of MeCP2 mutants. Percent of doubly transfected nuclei with TBL1X-mCherry/EGFP-MeCP2 spots in each independent transfection shown as mean (black line) and in- dividual values. WT n¼7, R306C, CTD1, CTD2 n¼4 independent transfections. Genotypes were compared to WT using a two-tailed Mann–Whitney test. R306C P¼0.0061 (**), CTD1 P¼0.0727 (ns), CTD2 P¼0.6485 (ns). (e) Representative western blot showing levels of MeCP2 protein present in neurons on day 15 of differentiation (7 days af- ter plating). Lanes A and B for each genotype are derived from independently targeted ES cell clones and WT pools are independent differentiations of the parental ES cell line. An antibody against the N-terminus of MeCP2 detects both full-length and truncated (D) protein. (f) Quantiﬁcation of MeCP2 levels in in vitro differentiated neurons. Two independent clones were differentiated 2 or 3 times for each genotype. MeCP2 level (MeCP2 signal/H3 signal) is normalized to the mean WT þ loxP value. Data shown are mean (black line) and ﬁve (WTþloxP, CTD1) or three (CTD2, WT pools) independent differentiations. Comparison to WTþloxP (two-tailed unpaired t-test with Welch’s correction for unequal variances): CTD1 P¼0.0021 (***), CTD2 P¼0.1874 (ns) and WT (pool) P¼0.4868 (ns). Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 Human Molecular Genetics, 2018, Vol. 27, No. 14 | 2539 ns (a)(b) 200 WT CTD1 CTD2 null MeCP2 MeCP2Δ H3 *** WT CTD1 CTD2 (c)(d) 10 100 CTD1 WT CTD1 8 80 null WT 6 60 null 4 40 2 20 0 0 4 8 12 16 20 24 28 32 0 4 8 12 16 20 24 28 32 36 Age (weeks) Weeks (e)(f) 60 CTD2 WT CTD2 WT 0 0 4 8 12 16 20 24 28 32 36 40 44 48 52 4 8 12 16 20 24 28 32 36 40 44 48 52 Age (weeks) Age (weeks) (g)(h)(i) Elevated Plus Maze Open Field (centre) Rotarod ns 100 ns 0.10 200 ns 80 0.08 0.06 ns 40 0.04 0.02 0 0.00 0 Closed Open CTD2 WT 123 Day CTD2 WT Figure 6. CTD1 and CTD2 mice show contrasting phenotypes due to different levels of MeCP2 in brain. (a) Western blot of whole brain lysates from 6-week-old CTD1, CTD2 and WT males. A Mecp2-null sample is shown for comparison. CTD1 and CTD2 mice express a truncated protein, MeCP2D.(b) Quantiﬁcation of (a): the level of MeCP2 (MeCP2 signal/H3 signal) as a percentage of mean WT shown as mean (black line) and individual values for each genotype (n¼3 animals). Comparison to WT (two-tailed unpaired t-test with Welch’s correction for unequal variances): CTD1 P¼0.0008 (***) and CTD2 P¼ 0.1773 (ns). (c) CTD1 hemizygous males display RTT-like phenotypes. Mean aggregate phenotypic score6SD plotted for mutants and WT male littermates, CTD1 n¼11, WT n¼7 (at 4 weeks). (d) Kaplan–Meier survival plot for CTD1 hemizygous males and WT littermates shown in (c). (e) Phenotypic scoring of CTD2 males and WT littermates. Mean aggregate phenotypic score6SD plotted for mutants and WT male littermates, CTD2 n¼16, WT n¼17. (f) Mean weight6SD for animals shown in (e). (g) Elevated plus maze. Percentage of time spent in the closed and open arms is shown as mean (black line) and individual values. Percentage time in open and closed arms for each genotype was compared using two-tailed un- paired t-tests: closed arms P¼0.2238 (ns), open arms P¼0.3713 (ns). (h) The ratio of distance travelled in the central zone of the open ﬁeld arena to total distance trav- elled shown as mean (black line) and individual values. Two-tailed unpaired t-test test P¼0.9394 (ns). (i) Accelerating rotarod test. The latency to fall is shown as mean (black line) and individual data points on each day. Two-way ANOVA (repeated measures) [genotype effect, F(1, 17)¼9.45110 , P¼ 0.9976 (ns)]. (g)–(i) CTD2 n¼9, WT n¼10. causal missense mutations disrupt the interactions with meth- Discussion ylated DNA or with the NCoR/SMRT co-repressor complexes Biochemical and cell biological studies of MeCP2 have identiﬁed (11). We were intrigued by a number of RTT mutations that numerous interacting proteins that potentially mediate its leave the MBD and the NID intact, and by the possibility that function. So far, only two of these are implicated in RTT, as they might direct us to additional functional domains that had Aggregate Score % time in arms Aggregate Score Distance: centre/total Weight (g) Percent survival Latency to fall (s) % WT mean Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 2540 | Human Molecular Genetics, 2018, Vol. 27, No. 14 (a) (b)(c) (d) (e) Figure 7. A humanized CTD2 allele expresses low levels of truncated MeCP2. (a) The amino acid sequences of the extreme C-termini of CTD1, CTD2 and CTD2hu mouse alleles and their corresponding human RTT alleles. Missense/nonsense changes caused by the deletion-frameshift are in bold type. Residues where the mouse allele differs from human are shown in red. Changes between CTD2 and CTD2hu are underlined. (b) Western blot showing levels of MeCP2 protein present in neurons on day 15 of differentiation (7 days after plating). Lanes A, B (and C) for each genotype are derived from independent mouse ES cell clones. An antibody against the N-terminus of MeCP2 detects both full-length and truncated (D) protein. (c) Quantiﬁcation of MeCP2 levels in in vitro differentiated neurons. Two or three independent clones were differentiated for each genotype. MeCP2 level (MeCP2 signal/H3 signal) is normalized to the mean WTþloxP value. Data shown are mean (black line) and individual val- ues for three (WTþloxP, CTD2hu) or two (CTD1, CTD2) independent differentiations. Comparison to WTþloxP (two-tailed unpaired t-test with Welch’s correction for unequal variances): CTD1 P¼0.0108 (*), CTD2 P¼0.5450 (ns) and CTD2hu P¼0.0111 (*). (d) Western blot showing levels of MeCP2 protein present in LUHMES-derived hu- man neurons on day 9 after precursor plating. Differentiations of one MECP2-null clone, two WT pools, two unmodiﬁed WT clones and three CTD2 LUHMES clones are shown. An antibody against the N-terminus of MeCP2 detects both full-length and truncated (D) protein. (e) Quantiﬁcation of MeCP2 levels in LUHMES-derived neurons. Three independent differentiations were performed for each of the clones shown in (d). MeCP2 level (MeCP2 signal/H3 signal) is normalized to the WT pool value for each differentiation. Data shown are mean (black line) and individual values for three independent differentiations. Comparison to WT A (two-tailed unpaired t-test): WT B P¼0.0798 (ns), CTD2 A P¼0.0015 (***), CTD2 B P¼0.0039 (***), CTD2 C P¼0.0027 (***). previously escaped detection. To be certain that the mutations detected several times in classical RTT (7), but has subsequently are genuinely causative of the disorder, we conﬁned our atten- been classiﬁed as a relatively common and phenotypically be- tion to variants shown to be absent in parental DNA. The impor- nign population variant (7,8). By studying C-terminal trunca- tance of this criterion is illustrated by E397K, which has been tions and the missense mutations P225R and P322L, which are Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 Human Molecular Genetics, 2018, Vol. 27, No. 14 | 2541 recurrent RTT mutations remote from the two known interac- and nonsense mutations R168X (23) and R255X (32). Our results tion sites, we failed to establish new functional domains of show that MeCP2 deﬁciency due to protein destabilization alone MeCP2. Instead, our results can provide coherent molecular is a major underlying cause of RTT. Although we only tested explanations for the deleterious effects of these mutations by two CTD mutations, we suggest that severe MeCP2 destabiliza- considering MeCP2 as a recruiter of histone deacetylase com- tion is a mechanism that applies generally to this RTT mutation plexes to chromatin (29). Further evidence that it is the confor- category. Why mutations outside the key functional domains mation and/or stability of MeCP2 that is affected by these should confer instability is unknown. With the exception of the mutations rather than functional domains containing P225, MBD and NID domains, whose 3D conformation has been deter- P322 or the CTD region comes from a severely truncated mouse mined (12,13), most of the protein is thought to be unstructured knock-in allele, DNIC, which despite lacking all of these sites (40). Surprisingly, the instability of CTD2 in the context of hu- shows a very mild phenotype, distinct from that seen in RTT man MeCP2 was suppressed by changing three nearby amino mouse models (14). acids to match the mouse MeCP2 sequence. This serendipi- It is clear from studies of patient severity (18,33,36,37) and tously allowed us to determine that the deletion of the MeCP2 mouse models (5,6,24,31) that mutations in the MBD and the CTD beyond amino acid 388, thereby losing 100 amino acids NID, such as T158M and R306C, are somewhat less severe on av- from the C-terminus, has no deleterious phenotypic consequen- erage than complete absence of the protein. This suggests that ces. We conclude therefore that this mutation does not identify these mutant proteins retain some function, which could come a novel functional domain, but causes RTT in humans purely from a number of sources. T158M does not completely abolish due to drastic destabilization of the protein. binding to methylated DNA (24) and some residual binding of The dramatic difference in the stability of CTD2 conferred by R306C to NCoR components can be seen in this study (Fig. 4b; human versus mouse sequences indicates that subtle changes Supplementary Material, Fig. S6b). All MeCP2 proteins tested in protein primary structure can have dramatic consequences could bind to mSin3a (Supplementary Material, Fig. S6b). It has for MeCP2 abundance. It is also interesting that CTD alleles that been proposed that other domains such as AT-hooks (38,39) result in reduced protein levels also have less mRNA, although may also contribute low-level function of MeCP2 when the MBD the decrease is less severe. Further investigation of the precise and NID are mutated. cause of CTD instability will be a focus of future research. P225R leads to moderate destabilization of MeCP2, but it also Importantly, the ﬁnding that CTD mutations do not appear to interferes with its ability to recruit the co-repressor subunit affect MeCP2 function may offer therapeutic opportunities. A re- TBL1X to chromatin and to impose DNA methylation- cent study has demonstrated the use of proteasome inhibitors dependent transcriptional repression. Although P225 is in an to increase the level of unstable T158M MeCP2 (46). unstructured region of the protein, it has been reported that Alternatively, small molecules that bind to the mutant CTD and conformational changes occur on binding to DNA (40). P225R affect its conformation may improve stability and therefore pro- may hinder the ability of MeCP2 to form the required bridge be- vide clinical beneﬁt. tween chromatin and co-repressor complex. It is also possible that the same conformational change leads to instability of the Materials and Methods protein. The combination of these two partial defects in stability and function can explain the moderate RTT-like phenotype of Mecp2 mutant mouse alleles these mice. Importantly, a rare human population variant at the RTT mutations P225R, P322L, CTD1, CTD2 and CTD2hu were in- same amino acid, P225A, does not interfere with co-repressor troduced into a targeting vector (20)(Fig. 2a) containing WT recruitment. The strong correlation between the ability to re- Mecp2 sequences and a Cre-excisable selection cassette using cruit co-repressor to methylated DNA and phenotypes in mice the QuikChangeII XL Site-Directed Mutagenesis kit (Agilent and humans strengthens the notion that this is a key role of Technologies). Targeting vectors were linearized at the 3 end MeCP2. using NotI and electroporated into 129/Ola E14Tg2a mouse ES The other non-canonical missense mutation, P322L, severely cells (a gift from A. Smith, University of Edinburgh). Correctly destabilizes MeCP2 and this alone can account for the resulting targeted clones were identiﬁed by an initial PCR screen: severe phenotype. The C-terminal truncations CTD1 and CTD2hu also severely destabilize the protein in neurons, but do 0 0 forward primer: 5 -TCACCATAACCAGCCTGCTCGC-3 not detectably compromise several aspects of its function in cell 0 0 reverse primer: 5 -ATTCGATGACCTCGAGGATCCG-3 ) transfection-based assays. Emphasizing this point, CTD2 mice, followed by Southern blotting of candidate clones and sequenc- which lack 100 amino acids from the C-terminal end of MeCP2, ing of the mutation sites (Supplementary Material, Figs S1 and are viable, fertile and phenotypically normal in a series of be- S9). A number of clones that had recombined to include the havioral assays. We note that this truncation removes an ﬂoxed selection cassette but no RTT mutation were also veriﬁed activity-dependent phosphorylation site at S421 (S423 in human for use as controls (WT þ lox). MeCP2) that has previously been implicated in neuronal func- tion, including dendritic patterning and spine morphogenesis (41–44). The region has also been implicated in regulation of Differentiation of ES cells into neurons micro-RNA processing and dendritic growth (45). Although loss of these functional domains evidently does not make a detect- ES clones were transiently transfected with a pCAGGS-Cre ex- able contribution to the major RTT-like phenotypes in mice, we pression plasmid and clones that had deleted the Neo selection cannot exclude the possibility that subtle phenotypic conse- cassette were veriﬁed by PCR, Southern blotting and sequencing quences were not detected by our assays or that they are impor- before differentiation into neurons. Two independently targeted tant for higher functions present in humans but not in mice. clones were used for each genotype, and each clone was differ- MeCP2 instability has previously been noted as a result of entiated into neurons two or three times. Differentiation was mutations that also disrupt the functional domains of MeCP2, carried out using a 4-/4þ retinoic acid (RA) procedure as previ- including missense mutations T158A (31), T158M and R133C (24) ously described (47,48). Neural precursor cells were seeded onto Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 2542 | Human Molecular Genetics, 2018, Vol. 27, No. 14 5 2 6 cm dishes at a density of 1.510 cells/cm and harvested after TGCCACTGCTCCCACCCCACCAGCCCCCCTGAGCCTCAGGACTTG 7 days by scraping into phosphate-buffered saline (PBS), pellet- AGCAGCAGCATCTGCAAAGAAGAGAAGATGC) (Sigma) into ing the cells and snap freezing. C57BL6/J: CBA/CaOlaHsd F2 fertilized oocytes as previously de- scribed (51). Correctly mutated founder animals were identiﬁed by sequencing across the deletion site. One CTD1 hemizygous Generation of CTD2 LUHMES cell lines male founder was able to breed and female heterozygous off- The LUHMES cell line (ATTC CRL-2927) was obtained from ATCC spring were checked by sequencing and Southern blot and cultured according to the methods described in Scholz et al. (Supplementary Material, Fig. S10a and b) and used for further (49), with some minor alterations. All vessels were coated in breeding of the line. poly-L-ornithine and ﬁbronectin overnight at 37 C. Proliferating LUHMES cells were seeded at 210 cells/T75 ﬂask every 2 days. Phenotypic analysis For differentiation, 2.510 cells were seeded in a T75 ﬂask for the ﬁrst 2 days of the protocol and on day 2 cells were seeded at Cohorts of hemizygous mutant male mice and WT littermates 610 cells/10 cm dish. During differentiation a half-media were weighed and scored weekly for a range of RTT-like pheno- change was performed on day 6 and neurons were harvested types to give an aggregate score between 0 and 12 as previously for protein on day 9. described (20,25). Cohorts of at least eight mutants and eight To introduce a CTD2 knock-in mutation, LUHMES cells were WT littermates were scored for each genotype, with higher transfected with plasmid pSpCas9(BB)-2A-GFP (pX458) (a gift numbers being used where available to allow for any losses from Feng Zhang, Addgene plasmid #48138) containing the unrelated to the mutation, such as ﬁghting. Scoring was carried 0 0 sgRNA sequence 5 -TCCTCGGAGCTCTCGGGCTC-3 , and single- out blind to genotype and to previous scores. Animals that stranded oligodeoxynucleotide 5 CCATCACCACCACTCAGAGTC scored a maximum score of 2 for tremor, breathing or general CCCAAAGGCCCCCGTGCCACTGCTCCCACCCCTGCCGCCCTGAGC condition or which had lost 20% of their body weight had CCCAGGACTTGAGCAGCAGCGTCTGCAAAGAGGAGAAGATGCCC reached the severity limit of the experiment according to the AGAGGAGGCT3 (Sigma, desalted). Cells were transfected by Home Ofﬁce license and were humanely culled. These animals Nucleofection (Lonza) using a Basic Nucleofector kit for primary were counted as having ‘died’ for the purposes of survival data. neurons (VAPI-1003) and a Nucleofector II device, and ﬂuores- Animals of any genotype which were culled for reasons not cence activated cell sorting (FACS)-sorted to isolate clones as linked to the mutation, such as ﬁghting with cage mates, were previously described (35). removed from survival plots at that point (censored data). LUHMES clones were analyzed by sequencing of genomic P225R, P322L and CTD2 animals had been backcrossed to DNA and cDNA and by Southern blotting (Supplementary C57BL6/J for three generations and CTD1 animals only once. Material, Fig. S11d and e) to isolate clones that were either ho- Behavioral testing was carried out on cohorts of mice that mozygous for the CTD2 mutation, or heterozygous and express- had been backcrossed to C57BL6/J for four generations. Testing ing the CTD2 mutation from the active X-chromosome took place over a 9-day period in the order: day 1: elevated plus (LUHMES cells are female, XX). Three CTD2 clones (two homozy- maze (115 min trial), day 2: open ﬁeld test (120 min trial), day gous and one heterozygous) and two unmodiﬁed WT control 3: wire suspension test (3 trials of 30 sec separated by 15 min), clones were used for further analysis. day 6: accelerating rotarod habituation (5 min at 4 rpm), days 7– LUHMES clones were differentiated to mature neurons as 9: accelerating rotarod trials (4 trials per day separated by 1 h). previously described (49). A null clone (H4) was described previ- Tests were performed as previously described (24,25). ously (35). All clones and the parental cell line underwent three Hemizygous male mutant mice and WT littermates were tested independent differentiations and were harvested on day 9 of at an age appropriate to the development of symptoms for that differentiation by scraping into PBS, pelleting the cells and snap line: testing was started for P322L at 6 weeks, P225R 11 weeks freezing. and CTD2 18 weeks. A cohort size of 10 animals per genotype was chosen as the largest number of animals that could reason- Establishment of Mecp2 mutant mouse lines ably be tested in a single session. Mice were housed in mixed mutant and WT littermate groups. Mice were tested blind to ge- P225R, P322L and CTD2 mouse lines were generated by injection notype and in order of ID number, giving a random order of mu- of ES cells into mouse blastocysts using standard methods. tant and WT animals. Positions on the rotarod were assigned to Chimeric males were crossed with CMVCre deleter females (50) ensure equal representation of mutant and control animals at to remove the selection cassette. This and subsequent genera- each of the ﬁve positions. tions were bred by crossing Mecp2-mutant heterozygote females with C57BL/6J WT males. The genotypes of mutant lines were conﬁrmed by Southern blotting (Supplementary Material, Figs RNA preparation and qRT-PCR S2a and b and S10d) and sequencing. Routine genotyping was Total cellular RNA was prepared from in vitro differentiated neu- performed by PCR across the loxP ‘scar’ site remaining in the rons (7 days after plating, 10 cells) and mouse brain (half brain) mutant alleles after selection cassette removal (forward primer: 0 0 using TriReagent (Sigma). cDNA was prepared using a p5 5 -TGGTAAAGACCCATGTGACCCAAG-3 , reverse primer: 0 0 QuantiTect kit (Qiagen) and ampliﬁed in a qPCR reaction using p7 5 -GGCTTGCCACATGACAAGAC-3 , WT 416bp, WT þ loxP SensiMix SYBR and Fluoroscein Master Mix (Bioline) using pri- 558bp). 0 0 mers for Mecp2 (forward: 5 -ACCTTGCCTGAAGGTTGGAC-3 , re- The CTD1 mouse line was created by pronuclear injection of 0 0 Cas9 protein (Integrated DNA Technology), synthetic tracrRNA verse: 5 -GCAATCAATTCTACTTTAGAGCGAAAA-3 ) and control 0 0 and crRNA (target sequence ACCTGAGCCTGAGAGCTCTG) Cyclophilin A (forward: 5 -TCGAGCTCTGAGCACTGGAG-3 , re- 0 0 verse: 5 -CATTATGGCGTGTAAAGTCACCA-3 ). Samples were (Sigma) and an oligonucleotide repair template (GCACCATCATC ACCACCATCACTCAGAGTCCACAAAGGCCCCCA run in triplicate and the amount of Mecp2 cDNA calculated for Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 Human Molecular Genetics, 2018, Vol. 27, No. 14 | 2543 each biological replicate using the ‘double delta’ method after double-transfected cells with DAPI/EGFP/mCherry co-localized correction of C values using a standard curve. spots was then calculated for each genotype in each indepen- dent transfection. Protein extracts and western blotting In vitro differentiated ES cell-derived and LUHMES-derived neu- Co-immunoprecipitation assay rons and whole brain samples were prepared for western blot- EGFP-MeCP2 expression constructs were transfected into HEK ting as previously described (3). Extracts were run on TGX 4–20% 293FT cells (R70007, ThermoFisher) using Lipofectamine and gradient gels (BioRad) loading extract equivalent to 5 10 nu- cell pellets were harvested 24 h after transfection and snap fro- clei per well. Samples were run on duplicate gels and trans- zen. Cell pellets were treated with Benzonase (Sigma) in low salt ferred to nitrocellulose membrane overnight in the cold at 25 V. NE1 buffer (20 mM HEPES pH7.5, 10 mM NaCl, 1 mM MgCl 0.1% 2, Western blots were processed as described previously (3) using Triton-X100, 10 mM b-mercaptoethanol and protease inhibitors) the following antibodies: anti-MeCP2 (N-terminus): mouse to release chromatin bound proteins and then extracted with monoclonal Men-8 (Sigma), anti-NeuN: rabbit polyclonal ABN78 150 mM NaCl. Supernatant extracts were mixed with GFP- (Millipore) and anti-histone H3: rabbit polyclonal ab1791 Trap _A beads (Chromotek) and after washing beads were (Abcam). Western blots were developed with IR-dye secondary boiled in SDS-PAGE sample buffer and released proteins and in- antibodies (IRDye 800CW donkey anti-mouse, IRDye 680LT don- put extract samples run on 4–15% TGX gradient gels (BioRad) key anti-rabbit, LI-COR Biosciences) and scanned using a LI-COR and blotted as above. Co-immunoprecipitated proteins were Odyssey machine. Images were quantiﬁed using Image Studio detected with the following antibodies: anti-NCoR: rabbit poly- Lite software (LI-COR Biosciences). The ratio of NeuN: histone H3 signals for each lane was used to check for equal neuronal clonal A301–145A (Bethyl Laboratories), anti-TBL1X: rabbit poly- in vitro differentiation and clones that did not differentiate well clonal ab24548 (Abcam), anti-HDAC3: mouse monoclonal 3E11 were discarded. MeCP2 levels were normalized to the histone (Sigma), anti-mSin3a: rabbit polyclonal ab3479 (Abcam) and H3 signal for each lane to compare the amount of MeCP2/nu- anti-EGFP: mouse monoclonal Living Colors JL-8 (Clontech). cleus between samples. Methylation-dependent repression assay EGFP-tagged cDNA constructs and transfection assays WT and mutant mouse Mecp2 e1 cDNAs were cloned into the Mouse Mecp2 cDNA (e2 isoform) was cloned into the vector expression vector pcDNA3.1(þ) IRES GFP, a gift from Kathleen L pEGFP-C1 (Clontech) to create in-frame fusions with an N-termi- Collins, Addgene plasmid #51406 (52). The methylation- nal EGFP tag. Missense mutations were introduced using the dependent repression assay was performed as previously de- QuikChangeII XL Site-Directed Mutagenesis kit (Agilent scribed (6) by co-transfecting MeCP2 expression plasmid, Fireﬂy Technologies) and truncations by PCR amplifying the cDNA luciferase (Luc) construct pGL2-control (with or without SssI with a modiﬁed reverse primer before cloning into the expres- methylation) and Renilla luciferase (Ren) transfection control sion vector. A vector expressing mCherry-tagged mouse TBL1X /y / plasmid into a Mecp2 , Mbd2 mouse tail ﬁbroblast cell line was described previously (30). For localization of EGFP-MeCP2 (6) using Lipofectamine. Luciferase signal was measured using and co-localization of EGFP-MeCP2/TBL1X-mCherry constructs the Dual Luciferase Assay kit (Promega). The ratio Luc/Ren sig- were transfected into NIH-3T3 mouse ﬁbroblasts (93061524, nal was calculated for each well and the mean unmethylated: ECACC) growing on glass coverslips using Lipofectamine (Thermo Fisher). Cells were ﬁxed in 4% paraformaldehyde 48 h mean methylated (U/M) value calculated for each genotype. after transfection, stained with DAPI and mounted in ProLong Each combination was repeated in triplicate in each experiment Diamond mountant (Thermo Fisher). Images were captured us- and each MeCP2 genotype was independently transfected at ing a Leica SP5 confocal microscope with 63 objective. For least four times. To check for equal expression of MeCP2 pro- western blotting to test expression levels of EGFP-MeCP2 WT tein, total protein extract from triplicate wells transfected with and mutant proteins cells were transfected with EGFP-MeCP2 either unmethylated or methylated luciferase template and and TBL1X-mCherry expression constructs as above and har- each of the MeCP2 expression constructs was western blotted vested 48 h after transfection by trypsinization. Cell pellets and the level of MeCP2 (normalized to c-tubulin) determined. were prepared for western blotting as described above for Primary antibodies: anti-MeCP2 rabbit monoclonal D4F3 (Cell in vitro differentiated neurons. MeCP2 was detected using Signaling Technology) and anti-c-tubulin mouse monoclonal mouse monoclonal Men-8 (Sigma) and anti-c-tubulin mouse GTU-88 (Sigma). monoclonal GTU-88 (Sigma) was used as a loading control. For quantiﬁcation of TBL1X-mCherry co-localization with EGFP-MeCP2, 3–7 independent transfections were performed for Statistical Analysis each genotype, with each transfection experiment including All statistical analysis was performed using GraphPad Prism EGFP-MeCP2 WT (positive control) and EGFP-MeCP2 R306C (neg- (GraphPad Software). Datasets were compared using Student’s ative control). Fields of cells were selected for scoring based on t-tests (with or without Welch’s correction for unequal variance, the presence of an EGFP (MeCP2) signal in at least one cell in the as appropriate), apart from the TBL1X-mCherry co-localization ﬁeld, without observing the mCherry channel. Cells with both assay and wire suspension test that were analyzed using a non- EGFP and mCherry ﬂuorescence were then scored for the pres- parametric Mann–Whitney test due to the nature of the data. ence or absence of TBL1X-mCherry spots in the nucleus, co-lo- Rotarod data were tested using two-way repeated measures calized with EGFP-MeCP2 and the DAPI bright spots. In each ANOVA, followed by post-hoc testing for signiﬁcance on each transfection at least 15 cells on 4 coverslips (total 60 cells per ge- notype) were counted, blind to the genotype. The percentage of day using the Holm–Sidak method. Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 2544 | Human Molecular Genetics, 2018, Vol. 27, No. 14 Cummings, B.B. et al. (2016) Analysis of protein-coding ge- Study Approval netic variation in 60, 706 humans. Nature, 536, 285–291. All animal experiments were performed under a United 9. Ebert, D.H., Gabel, H.W., Robinson, N.D., Kastan, N.R., Hu, Kingdom Home Ofﬁce project license (PPL no. 60/4547). L.S., Cohen, S., Navarro, A.J., Lyst, M.J., Ekiert, R., Bird, A.P. et al. (2013) Activity-dependent phosphorylation of MeCP2 threonine 308 regulates interaction with NCoR. Nature, 499, Data Availability 341–345. Data from this work are available from the authors. 10. Heckman, L.D., Chahrour, M.H. and Zoghbi, H.Y. (2014) Rett-causing mutations reveal two domains critical for Supplementary Material MeCP2 function and for toxicity in MECP2 duplication syn- drome mice. Elife, 3, e02676. Supplementary Material is available at HMG online. 11. Lyst, M.J. and Bird, A. (2015) Rett syndrome: a complex disor- der with simple roots. Nat. Rev. Genet., 16, 261–275. 12. Ho, K.L., McNae, I.W., Schmiedeberg, L., Klose, R.J., Bird, A.P. Acknowledgements and Walkinshaw, M.D. (2008) MeCP2 binding to DNA We would like to thank Alan McClure for animal husbandry and depends upon hydration at methyl-CpG. Mol. Cell, 29, Matthew Lyst and Dirk-Jan Kleinjan for critical reading of the 525–531. manuscript. A.B. is a member of the Simons Initiative for the 13. Kruusvee, V., Lyst, M.J., Taylor, C., Tarnauskaite, Z., Bird, A.P. Developing Brain at the University of Edinburgh. and Cook, A.G. (2017) Structure of the MeCP2-TBLR1 complex reveals a molecular basis for Rett syndrome and related dis- Conﬂict of Interest statement. A.B. is a member of the Board of orders. Proc. Natl. Acad. Sci. U S A, 114, E3243–E3250. ArRETT, a company based in the USA with the goal of develop- 14. Tillotson, R., Selfridge, J., Koerner, M.V., Gadalla, K.K.E., Guy, ing therapies for Rett syndrome. J., De Sousa, D., Hector, R.D., Cobb, S.R. and Bird, A. (2017) Radically truncated MeCP2 rescues Rett syndrome-like neu- Funding rological defects. Nature, 550, 398–401. 15. Gabel, H.W., Kinde, B., Stroud, H., Gilbert, C.S., Harmin, D.A., This work was supported by Wellcome (Investigator Award Kastan, N.R., Hemberg, M., Ebert, D.H. and Greenberg, M.E. 107930, PhD Studentship 109090 to L.F.) and the Rett Syndrome (2015) Disruption of DNA-methylation-dependent long gene Research Trust. Funding to pay the Open Access publication repression in Rett syndrome. Nature, 522, 89–93. charges for this article was provided by Wellcome. 16. Kinde, B., Wu, D.Y., Greenberg, M.E. and Gabel, H.W. (2016) DNA methylation in the gene body inﬂuences References MeCP2-mediated gene repression. Proc. Natl. Acad. Sci. U S A, 1. Amir, R.E., Van den Veyver, I.B., Wan, M., Tran, C.Q., Francke, 113, 15114–15119. U. and Zoghbi, H.Y. (1999) Rett syndrome is caused by muta- 17. Lagger, S., Connelly, J.C., Schweikert, G., Webb, S., Selfridge, J., Ramsahoye, B.H., Yu, M., He, C., Sanguinetti, G., Sowers, tions in X-linked MECP2, encoding methyl-CpG-binding pro- L.C., Walkinshaw, M.D. and Bird, A. (2017) MeCP2 recognizes tein 2. Nat. Genet., 23, 185–188. 2. Lewis, J.D., Meehan, R.R., Henzel, W.J., Maurer-Fogy, I., cytosine methylated tri-nucleotide and di-nucleotide Jeppesen, P., Klein, F. and Bird, A. (1992) Puriﬁcation, se- sequences to tune transcription in the mammalian brain. quence, and cellular localization of a novel chromosomal PLoS Genet., 13, e1006793. protein that binds to methylated DNA. Cell, 69, 905–914. 18. Bebbington, A., Percy, A., Christodoulou, J., Ravine, D., Ho, G., Jacoby, P., Anderson, A., Pineda, M., Ben Zeev, B., Bahi- 3. Ross, P.D., Guy, J., Selfridge, J., Kamal, B., Bahey, N., Tanner, K.E., Gillingwater, T.H., Jones, R.A., Loughrey, C.M. and Buisson, N., Smeets, E. and Leonard, H. (2010) Updating the McCarroll, C.S. (2016) Exclusive expression of MeCP2 in the proﬁle of C-terminal MECP2 deletions in Rett syndrome. nervous system distinguishes between brain and peripheral J. Med. Genet., 47, 242–248. 19. Smeets, E., Terhal, P., Casaer, P., Peters, A., Midro, A., Rett syndrome-like phenotypes. Hum. Mol. Genet., 25, 4389–4404. Schollen, E., van Roozendaal, K., Moog, U., Matthijs, G., 4. Skene, P.J., Illingworth, R.S., Webb, S., Kerr, A.R., James, K.D., Herbergs, J. et al. (2005) Rett syndrome in females with CTS hot spot deletions: a disorder proﬁle. Am. J. Med. Genet. A, Turner, D.J., Andrews, R. and Bird, A.P. (2010) Neuronal MeCP2 is expressed at near histone-octamer levels and glob- 132A, 117–120. ally alters the chromatin state. Mol. Cell, 37, 457–468. 20. Guy, J., Gan, J., Selfridge, J., Cobb, S. and Bird, A. (2007) 5. Chen, R.Z., Akbarian, S., Tudor, M. and Jaenisch, R. (2001) Reversal of neurological defects in a mouse model of Rett Deﬁciency of methyl-CpG binding protein-2 in CNS neurons syndrome. Science, 315, 1143–1147. results in a Rett-like phenotype in mice. Nat. Genet., 27, 21. Katz, D.M., Berger-Sweeney, J.E., Eubanks, J.H., Justice, M.J., Neul, J.L., Pozzo-Miller, L., Blue, M.E., Christian, D., Crawley, 327–331. 6. Guy, J., Hendrich, B., Holmes, M., Martin, J.E. and Bird, A. J.N., Giustetto, M. et al. (2012) Preclinical research in Rett syn- (2001) A mouse Mecp2-null mutation causes neurological drome: setting the foundation for translational success. Dis. symptoms that mimic Rett syndrome. Nat. Genet., 27, Model. Mech., 5, 733–745. 322–326. 22. Robinson, L., Guy, J., McKay, L., Brockett, E., Spike, R.C., 7. Christodoulou, J., Grimm, A., Maher, T. and Bennetts, B. Selfridge, J., De Sousa, D., Merusi, C., Riedel, G., Bird, A. et al. (2003) RettBASE: the IRSA MECP2 variation database-a new (2012) Morphological and functional reversal of phenotypes mutation database in evolution. Hum. Mutat., 21, 466–472. in a mouse model of Rett syndrome. Brain, 135, 2699–2710. 8. Lek, M., Karczewski, K.J., Minikel, E.V., Samocha, K.E., Banks, 23. Wegener, E., Brendel, C., Fischer, A., Hulsmann, S., Gartner, J. E., Fennell, T., O’Donnell-Luria, A.H., Ware, J.S., Hill, A.J., and Huppke, P. (2014) Characterization of the MeCP2R168X Downloaded from https://academic.oup.com/hmg/article/27/14/2531/4989385 by DeepDyve user on 19 July 2022 Human Molecular Genetics, 2018, Vol. 27, No. 14 | 2545 knockin mouse model for Rett syndrome. PLoS One, 9, phenotype of common mutations in Rett syndrome. J. Med. e115444. Genet., 41, 25–30. 24. Brown, K., Selfridge, J., Lagger, S., Connelly, J., De Sousa, D., 38. Baker, S.A., Chen, L., Wilkins, A.D., Yu, P., Lichtarge, O. and Zoghbi, H.Y. (2013) An AT-hook domain in MeCP2 deter- Kerr, A., Webb, S., Guy, J., Merusi, C., Koerner, M.V. et al. (2016) The molecular basis of variable phenotypic severity mines the clinical course of Rett syndrome and related dis- orders. Cell, 152, 984–996. among common missense mutations causing Rett syn- 39. Klose, R.J., Sarraf, S.A., Schmiedeberg, L., McDermott, S.M., drome. Hum. Mol. Genet., 25, 558–570. Stancheva, I. and Bird, A.P. (2005) DNA binding selectivity of 25. Cheval, H., Guy, J., Merusi, C., De Sousa, D., Selfridge, J. and MeCP2 due to a requirement for A/T sequences adjacent to Bird, A. (2012) Postnatal inactivation reveals enhanced re- methyl-CpG. Mol. Cell, 19, 667–678. quirement for MeCP2 at distinct age windows. Hum. Mol. 40. Ghosh, R.P., Nikitina, T., Horowitz-Scherer, R.A., Gierasch, Genet., 21, 3806–3814. L.M., Uversky, V.N., Hite, K., Hansen, J.C. and Woodcock, C.L. 26. Kerr, B., Alvarez-Saavedra, M., Saez, M.A., Saona, A. and (2010) Unique physical properties and interactions of the Young, J.I. (2008) Defective body-weight regulation, motor domains of methylated DNA binding protein 2. Biochemistry, control and abnormal social interactions in Mecp2 hypo- 49, 4395–4410. morphic mice. Hum. Mol. Genet., 17, 1707–1717. 41. Deng, J.V., Wan, Y., Wang, X., Cohen, S., Wetsel, W.C., 27. Samaco, R.C., Fryer, J.D., Ren, J., Fyffe, S., Chao, H.T., Sun, Y., Greenberg, M.E., Kenny, P.J., Calakos, N. and West, A.E. (2014) Greer, J.J., Zoghbi, H.Y. and Neul, J.L. (2008) A partial loss of MeCP2 phosphorylation limits psychostimulant-induced be- function allele of methyl-CpG-binding protein 2 predicts a havioral and neuronal plasticity. J. Neurosci., 34, 4519–4527. human neurodevelopmental syndrome. Hum. Mol. Genet., 17, 42. Li, H., Zhong, X., Chau, K.F., Santistevan, N.J., Guo, W., Kong, 1718–1727. G., Li, X., Kadakia, M., Masliah, J., Chi, J. et al. (2014) Cell 28. Nan, X., Tate, P., Li, E. and Bird, A. (1996) DNA methylation cycle-linked MeCP2 phosphorylation modulates adult neu- speciﬁes chromosomal localization of MeCP2. Mol. Cell. Biol., rogenesis involving the Notch signalling pathway. Nat. 16, 414–421. Commun., 5, 5601. 29. Nan, X., Ng, H.H., Johnson, C.A., Laherty, C.D., Turner, B.M., 43. Li, H., Zhong, X., Chau, K.F., Williams, E.C. and Chang, Q. Eisenman, R.N. and Bird, A. (1998) Transcriptional repression (2011) Loss of activity-induced phosphorylation of MeCP2 by the methyl-CpG-binding protein MeCP2 involves a his- enhances synaptogenesis, LTP and spatial memory. Nat. tone deacetylase complex. Nature, 393, 386–389. Neurosci., 14, 1001–1008. 30. Lyst, M.J., Ekiert, R., Ebert, D.H., Merusi, C., Nowak, J., 44. Zhou, Z., Hong, E.J., Cohen, S., Zhao, W.N., Ho, H.Y., Schmidt, Selfridge, J., Guy, J., Kastan, N.R., Robinson, N.D., de Lima L., Chen, W.G., Lin, Y., Savner, E., Grifﬁth, E.C. et al. (2006) Alves, F. et al. (2013) Rett syndrome mutations abolish the in- Brain-speciﬁc phosphorylation of MeCP2 regulates teraction of MeCP2 with the NCoR/SMRT co-repressor. Nat. activity-dependent Bdnf transcription, dendritic growth, Neurosci., 16, 898–902. and spine maturation. Neuron, 52, 255–269. 31. Gofﬁn, D., Allen, M., Zhang, L., Amorim, M., Wang, I.T., 45. Cheng, T.L., Wang, Z., Liao, Q., Zhu, Y., Zhou, W.H., Xu, W. Reyes, A.R., Mercado-Berton, A., Ong, C., Cohen, S., Hu, L. and Qiu, Z. (2014) MeCP2 suppresses nuclear microRNA proc- et al. (2012) Rett syndrome mutation MeCP2 T158A disrupts essing and dendritic growth by regulating the DNA binding, protein stability and ERP responses. Nat. DGCR8/Drosha complex. Dev. Cell, 28, 547–560. Neurosci., 15, 274–283. 46. Lamonica, J.M., Kwon, D.Y., Gofﬁn, D., Fenik, P., Johnson, 32. Pitcher, M.R., Herrera, J.A., Bufﬁngton, S.A., Kochukov, M.Y., B.S., Cui, Y., Guo, H., Veasey, S. and Zhou, Z. (2017) Elevating Merritt, J.K., Fisher, A.R., Schanen, N.C., Costa-Mattioli, M. expression of MeCP2 T158M rescues DNA binding and Rett and Neul, J.L. (2015) Rett syndrome like phenotypes in the syndrome-like phenotypes. J. Clin. Invest., 127, 1889–1904. R255X Mecp2 mutant mouse are rescued by MECP2 trans- 47. Bain, G., Kitchens, D., Yao, M., Huettner, J.E. and Gottlieb, D.I. gene. Hum. Mol. Genet., 24, 2662–2672. (1995) Embryonic stem cells express neuronal properties 33. Cuddapah, V.A., Pillai, R.B., Shekar, K.V., Lane, J.B., Motil, K.J., in vitro. Dev. Biol., 168, 342–357. Skinner, S.A., Tarquinio, D.C., Glaze, D.G., McGwin, G., 48. Li, M., Pevny, L., Lovell-Badge, R. and Smith, A. (1998) Kaufmann, W.E. et al. (2014) Methyl-CpG-binding protein 2 Generation of puriﬁed neural precursors from embryonic (MECP2) mutation type is associated with disease severity in stem cells by lineage selection. Curr. Biol., 8, 971–974. Rett syndrome. J. Med. Genet., 51, 152–158. 49. Scholz, D., Poltl, D., Genewsky, A., Weng, M., Waldmann, T., 34. Lotharius, J., Barg, S., Wiekop, P., Lundberg, C., Raymon, H.K. Schildknecht, S. and Leist, M. (2011) Rapid, complete and and Brundin, P. (2002) Effect of mutant alpha-synuclein on large-scale generation of post-mitotic neurons from the hu- dopamine homeostasis in a new human mesencephalic cell man LUHMES cell line. J. Neurochem., 119, 957–971. line. J. Biol. Chem., 277, 38884–38894. 50. Schwenk, F., Baron, U. and Rajewsky, K. (1995) A 35. Shah, R.R., Cholewa-Waclaw, J., Davies, F.C.J., Paton, K.M., cre-transgenic mouse strain for the ubiquitous deletion of Chaligne, R., Heard, E., Abbott, C.M. and Bird, A.P. (2016) loxP-ﬂanked gene segments including deletion in germ cells. Efﬁcient and versatile CRISPR engineering of human neu- Nucleic Acids Res., 23, 5080–5081. rons in culture to model neurological disorders. Wellcome 51. Aida, T., Chiyo, K., Usami, T., Ishikubo, H., Imahashi, R., Open Res., 1, 13. Wada, Y., Tanaka, K.F., Sakuma, T., Yamamoto, T. and 36. Bebbington, A., Anderson, A., Ravine, D., Fyfe, S., Pineda, M., Tanaka, K. (2015) Cloning-free CRISPR/Cas system facilitates de Klerk, N., Ben-Zeev, B., Yatawara, N., Percy, A., functional cassette knock-in in mice. Genome Biol., 16, 87. Kaufmann, W.E. et al. (2008) Investigating 52. Schaefer, M.R., Williams, M., Kulpa, D.A., Blakely, P.K., genotype-phenotype relationships in Rett syndrome using Yaffee, A.Q. and Collins, K.L. (2008) A novel trafﬁcking signal an international data set. Neurology, 70, 868–875. within the HLA-C cytoplasmic tail allows regulated expres- 37. Colvin, L., Leonard, H., de Klerk, N., Davis, M., Weaving, L., sion upon differentiation of macrophages. J. Immunol., 180, Williamson, S. and Christodoulou, J. (2004) Reﬁning the 7804–7817.

Journal

Human Molecular Genetics – Oxford University Press

Published: Jul 15, 2018

Keywords: mutation; neurons; mice; rett's disorder; phenotype; transcriptional repression

References

Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2

Amir, R.E.; Van den Veyver, I.B.; Wan, M.; Tran, C.Q.; Francke, U.; Zoghbi, H.Y.
Purification, sequence, and cellular localization of a novel chromosomal protein that binds to methylated DNA

Lewis, J.D.; Meehan, R.R.; Henzel, W.J.; Maurer-Fogy, I.; Jeppesen, P.; Klein, F.; Bird, A.
Exclusive expression of MeCP2 in the nervous system distinguishes between brain and peripheral Rett syndrome-like phenotypes

Ross, P.D.; Guy, J.; Selfridge, J.; Kamal, B.; Bahey, N.; Tanner, K.E.; Gillingwater, T.H.; Jones, R.A.; Loughrey, C.M.; McCarroll, C.S.
Neuronal MeCP2 is expressed at near histone-octamer levels and globally alters the chromatin state

Skene, P.J.; Illingworth, R.S.; Webb, S.; Kerr, A.R.; James, K.D.; Turner, D.J.; Andrews, R.; Bird, A.P.
Deficiency of methyl-CpG binding protein-2 in CNS neurons results in a Rett-like phenotype in mice

Chen, R.Z.; Akbarian, S.; Tudor, M.; Jaenisch, R.
A mouse Mecp2-null mutation causes neurological symptoms that mimic Rett syndrome

Guy, J.; Hendrich, B.; Holmes, M.; Martin, J.E.; Bird, A.
RettBASE: the IRSA MECP2 variation database-a new mutation database in evolution

Christodoulou, J.; Grimm, A.; Maher, T.; Bennetts, B.
Analysis of protein-coding genetic variation in 60, 706 humans

Lek, M.; Karczewski, K.J.; Minikel, E.V.; Samocha, K.E.; Banks, E.; Fennell, T.; O’Donnell-Luria, A.H.; Ware, J.S.; Hill, A.J.; Cummings, B.B.
Activity-dependent phosphorylation of MeCP2 threonine 308 regulates interaction with NCoR

Ebert, D.H.; Gabel, H.W.; Robinson, N.D.; Kastan, N.R.; Hu, L.S.; Cohen, S.; Navarro, A.J.; Lyst, M.J.; Ekiert, R.; Bird, A.P.
Rett-causing mutations reveal two domains critical for MeCP2 function and for toxicity in MECP2 duplication syndrome mice

Heckman, L.D.; Chahrour, M.H.; Zoghbi, H.Y.
Rett syndrome: a complex disorder with simple roots

Lyst, M.J.; Bird, A.
MeCP2 binding to DNA depends upon hydration at methyl-CpG

Ho, K.L.; McNae, I.W.; Schmiedeberg, L.; Klose, R.J.; Bird, A.P.; Walkinshaw, M.D.
Structure of the MeCP2-TBLR1 complex reveals a molecular basis for Rett syndrome and related disorders

Kruusvee, V.; Lyst, M.J.; Taylor, C.; Tarnauskaitė, Ž.; Bird, A.P.; Cook, A.G.
Radically truncated MeCP2 rescues Rett syndrome-like neurological defects

Tillotson, R.; Selfridge, J.; Koerner, M.V.; Gadalla, K.K.E.; Guy, J.; De Sousa, D.; Hector, R.D.; Cobb, S.R.; Bird, A.
Disruption of DNA-methylation-dependent long gene repression in Rett syndrome

Gabel, H.W.; Kinde, B.; Stroud, H.; Gilbert, C.S.; Harmin, D.A.; Kastan, N.R.; Hemberg, M.; Ebert, D.H.; Greenberg, M.E.
DNA methylation in the gene body influences MeCP2-mediated gene repression

Kinde, B.; Wu, D.Y.; Greenberg, M.E.; Gabel, H.W.
MeCP2 recognizes cytosine methylated tri-nucleotide and di-nucleotide sequences to tune transcription in the mammalian brain

Lagger, S.; Connelly, J.C.; Schweikert, G.; Webb, S.; Selfridge, J.; Ramsahoye, B.H.; Yu, M.; He, C.; Sanguinetti, G.; Sowers, L.C.; Walkinshaw, M.D.; Bird, A.
Updating the profile of C-terminal MECP2 deletions in Rett syndrome

Bebbington, A.; Percy, A.; Christodoulou, J.; Ravine, D.; Ho, G.; Jacoby, P.; Anderson, A.; Pineda, M.; Ben, Zeev B.; Bahi-Buisson, N.; Smeets, E.; Leonard, H.
Rett syndrome in females with CTS hot spot deletions: a disorder profile

Smeets, E.; Terhal, P.; Casaer, P.; Peters, A.; Midro, A.; Schollen, E.; van Roozendaal, K.; Moog, U.; Matthijs, G.; Herbergs, J.
Reversal of neurological defects in a mouse model of Rett syndrome

Guy, J.; Gan, J.; Selfridge, J.; Cobb, S.; Bird, A.
Preclinical research in Rett syndrome: setting the foundation for translational success

Katz, D.M.; Berger-Sweeney, J.E.; Eubanks, J.H.; Justice, M.J.; Neul, J.L.; Pozzo-Miller, L.; Blue, M.E.; Christian, D.; Crawley, J.N.; Giustetto, M.
Morphological and functional reversal of phenotypes in a mouse model of Rett syndrome

Robinson, L.; Guy, J.; McKay, L.; Brockett, E.; Spike, R.C.; Selfridge, J.; De Sousa, D.; Merusi, C.; Riedel, G.; Bird, A.
Characterization of the MeCP2R168X knockin mouse model for Rett syndrome

Wegener, E.; Brendel, C.; Fischer, A.; Hulsmann, S.; Gartner, J.; Huppke, P.
The molecular basis of variable phenotypic severity among common missense mutations causing Rett syndrome

Brown, K.; Selfridge, J.; Lagger, S.; Connelly, J.; De Sousa, D.; Kerr, A.; Webb, S.; Guy, J.; Merusi, C.; Koerner, M.V.
Postnatal inactivation reveals enhanced requirement for MeCP2 at distinct age windows

Cheval, H.; Guy, J.; Merusi, C.; De Sousa, D.; Selfridge, J.; Bird, A.
Defective body-weight regulation, motor control and abnormal social interactions in Mecp2 hypomorphic mice

Kerr, B.; Alvarez-Saavedra, M.; Saez, M.A.; Saona, A.; Young, J.I.
A partial loss of function allele of methyl-CpG-binding protein 2 predicts a human neurodevelopmental syndrome

Samaco, R.C.; Fryer, J.D.; Ren, J.; Fyffe, S.; Chao, H.T.; Sun, Y.; Greer, J.J.; Zoghbi, H.Y.; Neul, J.L.
DNA methylation specifies chromosomal localization of MeCP2

Nan, X.; Tate, P.; Li, E.; Bird, A.
Transcriptional repression by the methyl-CpG-binding protein MeCP2 involves a histone deacetylase complex

Nan, X.; Ng, H.H.; Johnson, C.A.; Laherty, C.D.; Turner, B.M.; Eisenman, R.N.; Bird, A.
Rett syndrome mutations abolish the interaction of MeCP2 with the NCoR/SMRT co-repressor

Lyst, M.J.; Ekiert, R.; Ebert, D.H.; Merusi, C.; Nowak, J.; Selfridge, J.; Guy, J.; Kastan, N.R.; Robinson, N.D.; de Lima, Alves F.
Rett syndrome mutation MeCP2 T158A disrupts DNA binding, protein stability and ERP responses

Goffin, D.; Allen, M.; Zhang, L.; Amorim, M.; Wang, I.T.; Reyes, A.R.; Mercado-Berton, A.; Ong, C.; Cohen, S.; Hu, L.
Rett syndrome like phenotypes in the R255X Mecp2 mutant mouse are rescued by MECP2 transgene

Pitcher, M.R.; Herrera, J.A.; Buffington, S.A.; Kochukov, M.Y.; Merritt, J.K.; Fisher, A.R.; Schanen, N.C.; Costa-Mattioli, M.; Neul, J.L.
Methyl-CpG-binding protein 2 (MECP2) mutation type is associated with disease severity in Rett syndrome

Cuddapah, V.A.; Pillai, R.B.; Shekar, K.V.; Lane, J.B.; Motil, K.J.; Skinner, S.A.; Tarquinio, D.C.; Glaze, D.G.; McGwin, G.; Kaufmann, W.E.
Effect of mutant alpha-synuclein on dopamine homeostasis in a new human mesencephalic cell line

Lotharius, J.; Barg, S.; Wiekop, P.; Lundberg, C.; Raymon, H.K.; Brundin, P.
Efficient and versatile CRISPR engineering of human neurons in culture to model neurological disorders

Shah, R.R.; Cholewa-Waclaw, J.; Davies, F.C.J.; Paton, K.M.; Chaligne, R.; Heard, E.; Abbott, C.M.; Bird, A.P.
Investigating genotype-phenotype relationships in Rett syndrome using an international data set

Bebbington, A.; Anderson, A.; Ravine, D.; Fyfe, S.; Pineda, M.; de Klerk, N.; Ben-Zeev, B.; Yatawara, N.; Percy, A.; Kaufmann, W.E.
Refining the phenotype of common mutations in Rett syndrome

Colvin, L.; Leonard, H.; de Klerk, N.; Davis, M.; Weaving, L.; Williamson, S.; Christodoulou, J.
An AT-hook domain in MeCP2 determines the clinical course of Rett syndrome and related disorders

Baker, S.A.; Chen, L.; Wilkins, A.D.; Yu, P.; Lichtarge, O.; Zoghbi, H.Y.
DNA binding selectivity of MeCP2 due to a requirement for A/T sequences adjacent to methyl-CpG

Klose, R.J.; Sarraf, S.A.; Schmiedeberg, L.; McDermott, S.M.; Stancheva, I.; Bird, A.P.
Unique physical properties and interactions of the domains of methylated DNA binding protein 2

Ghosh, R.P.; Nikitina, T.; Horowitz-Scherer, R.A.; Gierasch, L.M.; Uversky, V.N.; Hite, K.; Hansen, J.C.; Woodcock, C.L.
MeCP2 phosphorylation limits psychostimulant-induced behavioral and neuronal plasticity

Deng, J.V.; Wan, Y.; Wang, X.; Cohen, S.; Wetsel, W.C.; Greenberg, M.E.; Kenny, P.J.; Calakos, N.; West, A.E.
Cell cycle-linked MeCP2 phosphorylation modulates adult neurogenesis involving the Notch signalling pathway

Li, H.; Zhong, X.; Chau, K.F.; Santistevan, N.J.; Guo, W.; Kong, G.; Li, X.; Kadakia, M.; Masliah, J.; Chi, J.
Loss of activity-induced phosphorylation of MeCP2 enhances synaptogenesis, LTP and spatial memory

Li, H.; Zhong, X.; Chau, K.F.; Williams, E.C.; Chang, Q.
Brain-specific phosphorylation of MeCP2 regulates activity-dependent Bdnf transcription, dendritic growth, and spine maturation

Zhou, Z.; Hong, E.J.; Cohen, S.; Zhao, W.N.; Ho, H.Y.; Schmidt, L.; Chen, W.G.; Lin, Y.; Savner, E.; Griffith, E.C.
MeCP2 suppresses nuclear microRNA processing and dendritic growth by regulating the DGCR8/Drosha complex

Cheng, T.L.; Wang, Z.; Liao, Q.; Zhu, Y.; Zhou, W.H.; Xu, W.; Qiu, Z.
Elevating expression of MeCP2 T158M rescues DNA binding and Rett syndrome-like phenotypes

Lamonica, J.M.; Kwon, D.Y.; Goffin, D.; Fenik, P.; Johnson, B.S.; Cui, Y.; Guo, H.; Veasey, S.; Zhou, Z.
Embryonic stem cells express neuronal properties in vitro

Bain, G.; Kitchens, D.; Yao, M.; Huettner, J.E.; Gottlieb, D.I.
Generation of purified neural precursors from embryonic stem cells by lineage selection

Li, M.; Pevny, L.; Lovell-Badge, R.; Smith, A.
Rapid, complete and large-scale generation of post-mitotic neurons from the human LUHMES cell line

Scholz, D.; Poltl, D.; Genewsky, A.; Weng, M.; Waldmann, T.; Schildknecht, S.; Leist, M.
A cre-transgenic mouse strain for the ubiquitous deletion of loxP-flanked gene segments including deletion in germ cells

Schwenk, F.; Baron, U.; Rajewsky, K.
Cloning-free CRISPR/Cas system facilitates functional cassette knock-in in mice

Aida, T.; Chiyo, K.; Usami, T.; Ishikubo, H.; Imahashi, R.; Wada, Y.; Tanaka, K.F.; Sakuma, T.; Yamamoto, T.; Tanaka, K.
A novel trafficking signal within the HLA-C cytoplasmic tail allows regulated expression upon differentiation of macrophages

Schaefer, M.R.; Williams, M.; Kulpa, D.A.; Blakely, P.K.; Yaffee, A.Q.; Collins, K.L.

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

A mutation-led search for novel functional domains in MeCP2

A mutation-led search for novel functional domains in MeCP2

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

A mutation-led search for novel functional domains in MeCP2

A mutation-led search for novel functional domains in MeCP2

Abstract

Journal

Recommended Articles

References

Our policy towards the use of cookies