Results

Exome analysis of case-parent trios with cardiovascular laterality defects

As outlined above, the results of exome sequencing of 14 case-parent trios were already published by Breuer et al. in 202211. In two further case-parent trios (HET15 and HET17) we found an identical de novo ultra-rare missense variant in the FAM222B gene (Fig. 1A, cases 1&2). This variant has been identified applying the same filter criteria as previously described by Breuer et al.11. The FAM222B variant SNV: 17–28,759,060-G-A (GRCh38, dbSNP rs753661690, c.899G > A, p.Arg300His) has been reported in only 20 of 1.612.112 alleles (minor allele frequency, MAF 0.00001241) without reports of homozygotes (gnomAD v4.1.0). The CADD score is 24.4. Both affected individuals presented with left isomerism and complex CHD. Screening of homozygous variants in HET15 with consanguineous marriage did not yield any hits.

Fig. 1: Patient pedigrees and genomic comparison of zf and human FAM222B. (A) Pedigrees of cases with cardiovascular laterality defects, situs inversus totalis, heterotaxy, or isolated CHD identified through exome sequencing. Circle for female, square for male. Doubled line for consanguineous marriage. Affected patients in black: I1: lifelong reduced physical capacity, late diagnosis of an unspecified ASD; II2: PDA, patent foramen ovale (PFO); II3: congenital valvular pulmonary stenosis; III1: VSD, PDA, ASD II, bicuspid aortic valve with stenosis and insufficiency; III2: multiple VSDs, double chambered right ventricle, PDA, bicuspid aortic valve with insufficiency; III3: VSD, PDA, PFO; III4: common truncus arteriosus type A1, ASD II, left superior vena cava. * = No DNA available (B) Variant table of FAM222B including specific changes in base and amino acid sequence, MAF, CADD and ACMG, VUS = variant of uncertain significance, * cannot be classified according to ACMG criteria. (C) Genetic structure of human FAM222B, the zf homologues fam222ba/bb and zf fam222aa. The introns (lines), exons (boxes) and protein coding (dark blue) sequence sizes were adjusted according to the entire gene. Dark green bands underneath the human FAM222B reveal the genetic conservation score compared to zf (UCSC genome browser). The six variants identified in our studies are marked with an arrow (point-mutations) or line (deletion) in the highly conserved exon three. Variant 1/2, which was further observed in our studies, was highlighted in red; other variants were highlighted in orange. Red bars in the zf paralogues reveal the position of the 10 bp deletion in the TALEN-generated fam222ba/bb dd-KO line, leading to a reading frame shift with a resulting early stop codon. Confirmative sequencing of the received dd-KO zfl is shown in the box aside. Blue numbers beneath each chromosome label represent a comparative analysis between human FAM222B and zf homologues Fam222ba, Fam222bb as well as zf Fam222aa on amino acid level using the UniProt align tool. The highest similarity was revealed between FAM222B and Fam222bb (59.9%), followed by Fam222ba (54.2%) and Fam222aa (34.6%).

Exome survey of 2,109 single exomes in individuals with situs inversus totalis, heterotaxy, or isolated CHD

Using the same filter criteria as applied for our unbiased exome survey in case-parent trios, we surveyed 2,109 single exomes of individuals with situs inversus totalis, heterotaxy, or isolated CHD and identified five additional variants in FAM222B in five independent families (Fig. 1A, cases 3–7; Fig. 1B). The five variants comprised four heterozygous missense variants and one in-frame deletion of seven amino acids followed by one amino acid insertion. The affected variant carriers presented with isolated heart defects such as atrial septal defect (ASD), ventricular septal defect (VSD), atrioventricular septal defect (AVSD) and bicuspid aortic valve. More detailed filtering of exome data in these five families did not identify another plausible variant in previously established disease genes associated with congenital heart defects.

In individual “III-1, EGAN00001389505”, who presented with congenital VSD, patent ductus arteriosus (PDA), ASD II and bicuspid aortic valve with stenosis and insufficiency, we found a missense variant c.485A > T (p.Gln162Leu, NM_001077498.3, rs768625209) with a CADD score of 23.8 (Fig. 1A, case 3). The variant has been reported only once in 1.563.132 alleles (MAF 0.0000006397) without reports of homozygotes (gnomAD v4.1.0). Within a three-generation multiplex family, the variant segregated between all seven differently affected family members. All PDAs occurring in this family were treated with interventional catheter or were surgically ligated in early infancy.

In individual “EGAN00001377228”, who presented with congenital ASD, we found a missense variant c.316A > G (p.106Lys > Glu, NM_001077498.3, rs1247152952) with a CADD score of 25.4 (Fig. 1A, case 4). The variant has been reported in 22 of 1.613.164 alleles (MAF 0.00001364) without reports of homozygotes (gnomAD v4.1.0). Genetic material of the father was not available.

In individual “EGAN00001389854”, who presented with congenital bicuspid aortic valve, we found a missense variant c.419C > T (p.140Ala > Val, NM_001077498.3) with a CADD score of 23.5 (Fig. 1A, case 5). The variant has been reported in only 2 of 1.600.126 alleles (MAF 0.000001250) without reports of homozygotes (gnomAD v4.1.0). Genetic material of the father was not available.

In individual “EGAN00001073428”, who presented with congenital AVSD, we found an in-frame deletion of seven amino acids followed by one amino acid insertion c.558_575delGACGACCTCCGAGACTCCC > G (p.186_192delHisProGlnSerLeuGlnGln > Gln, NM_001077498.3) in a conserved region of FAM222B (Fig. 1A, case 6). This deletion has not been previously described. Parental DNA was not available.

In individual EGAN00001377174, who presented with congenital ASD, we found a missense variant c.1148C > T (p.383Thr > Met, NM_001077498.3, rs758359508) with a CADD score of 22.7 (Fig. 1A, case 7). The variant has been reported in 27 of 1.581.696 alleles (MAF 0.00001707) without reports of homozygotes (gnomAD v4.1.0). Parental DNA was not available.

All missense variants reported in our study have been previously reported in gnomAD with a MAF ranging from < 0.00001707 to 0.0000006397 compared to a MAF of 0.000235 for all missense variants in families 3, 4, 5 and 7 or 0.000470 for families 1 and 2 in the present sample here (combined cohort of discovery sample and single exomes). The frequencies of the missense variants identified in families 1, 2, 3, and 5 were significantly higher in the present case cohort than in the general population. In contrast, in families 4 and 7, the identified variant occurred at a comparable frequency in the combined cohort and the general population. The nonsense indel variant identified in family 6 is not represented in the gnomAD at all (Fig. 1B).

Using the UCSC genome browser we found all of our patients ‘ FAM222B variants localized in the highly conserved exon 3 region (Fig. 1C).

Functional analysis of FAM222B in the zebrafish larvae model

For a better understanding of our candidate gene function, we chose the zfl (danio rerio) as an in vivo development model. To do so, we analysed the gene structure of human FAM222B and compared it to the zf fam222b homologues and fam222aa (Fig. 1C). We further reviewed the conservation of human FAM222B compared to zf using the UniProt align tool. The highest similarity on amino acid level between human FAM222B and zf was observed with 59.9% for Fam222bb. This was followed by Fam222ba with 54.2% and Fam222aa with 34.6% (Fig. 1C). Comparative analysis of the amino acid conservation of the observed human variant gene-loci in zf revealed the highest conservation in zf Fam222bb with three out of six human variant loci preserved here (see Supplementary S1 online). Notably, the gene-locus of our human variant of interest 1/2, rs753661690, p.300Arg > His is preserved in zf Fam222bb as well as in zf Fam222ba, prompting this variant as best suitable for our further analysis. Genomic comparison of up- and downstream neighbouring genes of human FAM222B, with the zf’ homologues fam222ba/bb and fam222aa shows the highest conservation in this case with fam222ba followed by fam222bb and no common gene neighbours with fam222aa (see Supplementary S2 online).

Control sequencing of the obtained TALEN-derived Tg(kdrl:eGFP; gata ds:red) fam222bb hu10539; fam222ba hu10755; s843Tg (ZDB-FISH-161130–1) dd-KO strain is concordant to the given 10 bp deletion from our collaborators and zfin (Fig. 1C).

Whole-mount in situ hybridization (WISH) of fam222ba/bb in zf cardiac development

For analysis of fam222ba, fam222bb and fam222aa mRNA expression patterns in wt and fam222ba/bb dd-KO zfl, we performed whole-mount in situ hybridization throughout various time-points between 12 and 72 hpf (see Supplementary S3 online). Due to highest overall expression patterns in 48 hpf old zfl, further investigations were performed at this time-point (Fig. 2A). Expression patterns of all three fam222 homologues were restricted to the head and the cardiac region in a similar pattern (Fig. 2A). There were no obvious expression differences in any of the fam222 homologue stainings for the fam222ba/bb dd-KO (Fig. 2A: c,f,i) compared to wt zfl (Fig. 2A: a,d,g). To closer proof expression of fam222 especially in the developing zfl heart, double staining against ADP-Ribosylhydrolase Like 1 enzyme (adprhl1) as a cardiac marker was performed in wt, displaying a clear overlay of both probes in the cardiac region, supporting the role of fam222ba/bb as a regulator of cardiac development (Fig. 2A: b,e,h). WISH analysis at different time-points revealed additional staining in the cloaca region for fam222ba at 24 hpf and in the pineal gland for fam222bb at 72 hpf (see Supplementary S3 online).

Fig. 2: Whole-mount in situ hybridization (WISH) analysis against fam222 homologues and cardiac markers adprhl1 and myl7 in wt and fam222ba/bb dd-KO zfl at 48 hpf. (A) Expression of fam222 homologues at 48 hpf using three different probes against each zf fam222 gene: fam222ba (a-c), fam222bb (d-f), fam222aa (g-i). Staining in wt zfl (a,d,g) compared to fam222ba/bb dd-KO zfl (c,f,i). Blue WISH staining is apparent in the zf head region as well as in the pericardium region including the heart. Additional red staining of the heart was performed using a probe against cardiac adprhl1 (j). Double staining was realized in wt zfl (b,e,h) in order to test for overlay of fam222 in blue and adprhl1 in red. Scale bar at the bottom applies for whole row above: 500 µm. (B) WISH-analysis to examine heart laterality defects. Myl7 indicates cardiac looping in wt and fam222ba/bb dd-KO zfl at 48 hpf. Analysis of various ways of cardiac looping: d-loop (right, normal), no loop (mid), l-loop (left, inversed). Scale bar 100 µm. (C) Direction and amount of false directed heart looping in wt and fam222ba/bb dd-KO zfl (N = 6, n = 148–173 in each group, in total), no loop in purple, l-loop in lavender. Two-sided Chi2 test: p-value: 0.0017, ** = very significant. (D) Comparative analysis of whole-mount Tg(kdrl:EGFP) wt and Tg(kdrl:EGFP) fam222ba/bb dd-KO zfl at 48 hpf showed no difference in fluorescent in vivo imaging concerning morphological and cardiovascular development between both groups. Lateral view of whole-mount zfl in bright field (a,d). (b,e) Lateral view of whole-mount zfl in fluorescence mode (b,e). Lateral view of zoomed in cranial head & heart region of whole-mount zfl in fluorescence mode (c,f). Abbreviations: asterisks* = heart, AT = atrium, V = ventricle. Scale bar for (a,b,d,e): 500 µm. Scale bar for (c,f): 200 µm. (E) Comparison of heartbeats per minute reveals no difference between wt and dd-KO zfl (N = 3, n = 30 in each group). Median (bpm): wt: 115, dd-KO: 116.5. Unpaired two-tailed t-test: p-value: 0.938, ns = not significant.

To further investigate the role of zf fam222ba/bb in heart development and looping, we performed a WISH-analysis at 48 hpf comparing wt and fam222ba/bb dd-KO zfl using an anti-myl7 probe to detect heart laterality defects (Fig. 2B). Our data strongly showed higher frequency of cardiac looping defects in dd-KO zf hearts compared to wt zfl. Whereas heart looping in wt mostly presented itself as normally d-looped (97.6%), the amount of properly d-looped hearts was significantly lower in fam222ba/bb dd-KO zfl (89.2%), p-value: < 0.01**. In both groups, wt and dd-KO, the distribution of l-looped and no looped hearts was equal (Fig. 2C). Based on these results, our data strongly supports fam222ba/bb dd-KO in modulating heart looping.

Comparative in vivo imaging analysis focusing on the morphological and cardiovascular development between whole-mount Tg(kdrl:EGFP) wt and Tg(kdrl:EGFP) fam222ba/bb dd-KO zfl at 48 hpf revealed no obvious difference between both groups (Fig. 2D). In addition to this, also the heart rate (in beats per minute, bpm, at 48 hpf of the zfl) showed no significant difference between the two groups wt and dd-KO (p-value: > 0.05, Fig. 2E). Moreover, no cardiac arrhythmias were observed in any group.

Cardiac analysis of fam222ba/bb knockout in adult zebrafish

Detailed examination and comparison of adult (≈ two-year-old) wt and fam222ba/bb dd-KO zf hearts showed structural anomalies in the anatomical composition and proportion of the dd-KO zf hearts. For analysis of proportion, area measurements were adjusted to individual zf weights. There was no significant difference between wt and fam222ba/bb dd-KO weights, p-value: > 0.05 (see Supplementary S4 online). In some dd-KO zf, the bulbus arteriosus (BA) seemed enlarged (Fig. 3A f), however, this phenotype did not appear in a significant difference in comparison to wt zf (p-value: > 0.05). One out of 16 analysed dd-KO zf presented itself with a doubled BA, no anomalies were observed in total 17 wt zf BA (see Supplementary S5 online).

Fig. 3: Examination of adult zf hearts, morphology and histology. (A) Detailed examination and comparison of dissected adult wt and fam222ba/bb dd-KO zf ‘ hearts (N = 4, n = 17 in each group, except for BA analysis in dd-KO zf group: N = 4, n = 16, in total): (a-c) show representative examples of normally developed wt zf hearts, (d-f) demonstrate various representative altered fam222ba/bb dd-KO zf hearts. (d) presents a hypertrophic atrium, (e) shows a slim ventricle and (f) reveals a hypertrophic BA. Scale bar: 1000 µm. Dashed lines symbolize the way of measurement exemplary in c&f for the zf ‘ ventricle in blue, the atrium in green and the BA in black. Abbreviations: AT = atrium, BA = bulbus arteriosus, V = ventricle. No significant difference between BA area of wt and fam222ba/bb dd-KO zf. Median (mm2/ mg): wt: 0.0010, dd-KO: 0.0011. Unpaired two-tailed t-test: p-value: 0.366, ns = not significant. (B) Box plot graph comparing wt and fam222ba/bb dd-KO zf ‘ ventricle area. Median (mm2/ mg): wt: 0.0022, dd-KO: 0.0027. Unpaired two-tailed t-test: p-value: 0.007, ** = very significant. (C) Box plot graph comparing wt and fam222ba/bb dd-KO zf ‘ atrium area. Median (mm2/ mg): wt: 0.0012, dd-KO: 0.0018. Unpaired two-tailed t-test: p-value: 0.011, * = significant. (D) HE-Staining of adult wt (a,b) and dd-KO (c,d) zf hearts. Each box in the overview (a,c) corresponds to a 50 µm × 50 µm area used for quantification of nuclei. Both areas enclosed by thicker (bold) boxes are presented examples at higher resolution in the corresponding magnified panel (b,d). Scale bar: overview: 500 µm, zoom: 10 µm. Abbreviations: AT = atrium, BA = bulbus arteriosus, V = ventricle. (E) Box plot graph comparing the number of nuclei in each selected wt and fam222ba/bb dd-KO heart in five 50 µm × 50 µm areas of ventricular myocardium (N = 4, n = 20). Median: wt: 48, dd-KO: 45. Unpaired two-tailed t-test: p-value: 0.030, * = significant. (F) TEM images showing ventricular myocardium of adult wt (a,b) and fam222ba/bb dd-KO (c,d) zf. Overview (a,c) and higher magnification (b,d) of sarcomeres. Box in overview indicates zoomed section. There was no apparent difference neither in the overall tissue assembly nor in the detailed composition of sarcomeres. Scale bar: overview: 5 µm, zoom: 500 nm.

Remarkably, adult dd-KO zf presented with enlarged hypertrophic ventricle and atrium area compared to adult wt zf in our 2D measurements (Fig. 3A,B, p-value: < 0.01**; Fig. 3A,C, p-value: < 0.05*).

Unexpectedly the amount of cell nuclei in adult dd-KO ventricular heart tissue was significantly decreased (Fig. 3D,E, p-value: < 0.05*). Cell density and tissue area were quantified in dd-KO and WT zf heart samples. Notably, for both groups analysis areas with approximately confluent cellular organization without detectable intercellular spaces were selected. The reduced cell density observed in dd-KO heart tissue compared to WT (44.63 vs. 47.95 cells per 2500 µm2) strongly suggests an increase in average cell size. This is consistent with a hypertrophic phenotype at the cellular level. In addition to that, the total tissue area normalized to organismal mass was higher in KO than in WT (0.002929 vs. 0.002422 mm2/mg). By integrating these metrics, the estimated number of cells per mg was calculated in our studies, revealing a higher overall cell number in KO compared to WT (0.1307 vs. 0.1161 cells/mg), corresponding to an increase of approximately 12.6%, which indicates cellular hyperplasia in dd-KO zf tissue.

Through Masson’s Trichrome Stain it was also ensured that no sample was biased by fibrous tissue infiltration. With the exception of the physiological fibrous heart valves, no fibrous tissue or scarring was observed. No differences were detected between KO and WT (see Supplementary S6 online).

Further histological differences were not obvious in our ultrastructural histological nor electron microscopy analysis (Fig. 3F), which do not reveal any significant difference between wt and dd-KO composition in closer muscle ultra-structure sarcomere assembly.

Testing of c.899G > A variant in embryonic cardiac development in zfl using mRNA injection

FAM222B encodes for an approximately 60 kDa nucleoplasmic and mitochondrial protein as annotated in the UniProt database. The protein does not contain any typical signal-peptide, transmembrane nor membrane-anchor regions12. Structure prediction by AlphaFold2 suggests an intrinsically unstructured protein with only five sparsely set helices within the 562 amino acids of the human variant. Through database research (STRING network) we could discover several potential interaction partners of FAM222B, including PDZD9, NLK, OR51D1, ZNF572, SSC4D, CEP152, OR2K2, ZNF789, TRMT10B and ZNF695 (listed according to scores). While these potential interaction partners were suggested from text mining, experimental evidence was only provided for NLK (STRING network: BioPlexExplorer, version BioPlex 3.020; IntAct Molecular Interaction Database, version: 1.0.421, both queries on 18th September 2025) which is also known to play an indirect role in cardiac development and function22. In a next step, we modelled protein complexes of FAM222B with the potential interaction partners, as many intrinsically unstructured proteins are known to adopt well-defined conformations only upon binding to their target proteins. Using AlphaFold323, we calculated FAM222B complex structures with PDZD9, NLK, ZNF572, SSC4D and CEP152, always as heterodimers and heterotetramers (dimers of dimers). Following this approach, only the heterodimeric interaction of FAM222B with NLK revealed convergent results, with one helical element of FAM222B interacting with the N-lobe of NLK, suggesting the protein of interest as a substrate of the kinase. Notably, Arg300 of FAM222B mediates an intense interaction with the β-sandwich region of the N-lobe in all calculated models, positioning the phosphorylation-site Ser296 in the catalytic centre of the kinase (Fig. 4A). Moreover, it becomes apparent that the patient-derived histidine variant His300, rs753661690 cannot participate in the hydrogen-bonding network formed by arginine due to the significantly shorter side chain (Fig. 4A‘). This may comprise kinase recognition, resulting in the loss of phosphorylation of FAM222B. Of note, a recent study supported the interaction of FAM222B with NLK, suggesting that an N-terminal intrinsically disordered region in FAM222 family proteins interacts with the backside of the kinase, thus mediating the recruitment of the substrate24. Following these modelling results and considering the appearance of our found SNV p.300Arg > His in two unrelated families (Fig. 1A, cases 1&2), we focused on this mutation side in the zf model for further functional analysis.

Fig. 4: Modelling and functional effects of the FAM222B variant 1/2 rs753661690. (A) Structural model of human FAM222B (residues 286–314) bound to human NLK (residues 140–425). Arg300 (R300) of FAM222B is interacting with the β-sandwich structure of the kinase N-lobe, placing the preceding Ser296 (S296) residue in a position facing the bound nucleotide. Shown is the calculated model where the co-substrate ATP is complexed with two magnesium ions, with the γ-phosphate facing the hydroxyl group of Ser296 of the FAM222B sequence segment S296PISR300. (A’) Close up of catalytic centre showing hydrogen bonds between Arg300 and NLK as well as variant situation showing His300 lacking these. (B) Representative bright field overview and fluorescent heart images of poly(A) mRNA injected wt zfl. After mRNA injection, the 48 hpf old zfl were classified into four different severities of morphological and accompanying cardiac phenotypes: normal, mild (e.g. pericardial edema), strong (e.g. hypotrophic ventricle and concomitant dilated/ hypertrophic atrium) and severe (e.g. no heart looping and no grip). Scale bar: bright field: 500 µm, fluorescent cardiac region: 100 µm. (C) Bar chart classifies zfl (%) into three groups according to the specific developmental disorder: only morphological development (grey) affected, combined cardiac and morphological disorder (grey with green stripes) and only cardiac development (green) affected. N = 3, n = 38–59 in each group, in total. Overall phenotype: two-sided Chi2 test: 150 ng/µl wt against 75 ng/µl variant: p-value: 0.0389, *; 150 ng/µl wt against 150 ng/µl variant: p-value: 0.0499, *. Only morphologically changed zfl: two-sided Chi2 test: 150 ng/µl wt against 75 ng/µl variant: p-value: 0.677, ns; 150 ng/µl wt against 150 ng/µl variant: 0.961, ns. Combined phenotype, including a topical morphological and a cardiac phenotype together in one zfl: two-sided Chi2 test: 150 ng/µl wt against 75 ng/µl variant: p-value: 0.063, ns; 150 ng/µl wt against 150 ng/µl variant: p-value: 0.306, ns. Only cardiac altered zf: two-sided Fisher’s exact test: 150 ng/µl wt against 75 ng/µl variant: p-value: 1, ns; 150 ng/µl wt against 150 ng/µl variant: p-value: 0.056, ns. (C’) Bar chart accumulates all heart phenotypes (combined morphological + cardiac disorder in grey box with green stripes and only cardiac phenotype in green box) and is separately illustrated, meaning taken from (C), to underline differences in the amount of pathologically altered hearts in each group.

Thus, we reconstructed this specific variant via mutagenesis of FAM222B wt DNA and investigated its effect on heart development performing poly(A) mRNA injection in Tg(kdrl:EGFP) wt zf embryos. For a clearer differentiation, we compared overexpression of human unmodified (wt) FAM222B mRNA to injection with the above-mentioned human missense variant rs753661690 SNV p.300Arg > His (var. 1/2) mRNA. We employed two different concentrations of mRNA for injections and compared the effects found on the zf ‘ morphological and cardiac phenotype to un-injected as well as embryos control-injected with phenol red and buffered water. After mRNA injection, the 48 hpf old zfl were classified into four different severities of morphological and cardiac phenotypes: normal, mild (e.g. pericardial edema), strong (e.g. hypotrophic ventricle and concomitant dilated/ hypertrophic atrium), and severe (e.g. no heart looping and no grip) (Fig. 4B). Statistical analysis was performed differentiating the distribution of all phenotypes (Fig. 4C).

Zf injected with FAM222B variant mRNA represent a higher percentage of phenotypes in total (75 ng/µl: 49.1%; 150 ng/µl: 50.0%) than zf injected with the overexpressed human wt FAM222B mRNA (150 ng/µl: 28.3%). This can also be found significant for both concentrations with a p-value of < 0.05* (Fig. 4C).

Whereas the number of only morphologically changed zf is approximately similar in each group: wt FAM222B 150 ng/µl: 15.4%; FAM222B rs753661690 (var. 1/2) 75 ng/µl: 18.6%; FAM222B rs753661690 (var. 1/2) 150 ng/µl: 15.8%, p-value: > 0.05, (Fig. 4C), the number of zfl with heart phenotypes in each group differs strongly:

Of note, the injection of FAM222B variant rs753661690 (var. 1/2) shows a cascading effect on the presentation of a cardiac phenotype. While the wt FAM222B 150 ng/µl group shows an isolated cardiac phenotype in only 2.6% of cases, this percentage increases in a dose-dependent manner to 5.1% in FAM222B rs753661690 (var. 1/2) 75 ng/µl and even 15.8% in FAM222B rs753661690 (var. 1/2) 150 ng/µl (p-value: = 0.0564) (Fig. 4C).

For a combined phenotype, including a morphological and a cardiac phenotype together in one zfl, it becomes apparent that an overexpression of wt human FAM222B mRNA led, with a phenotype percentage of 10.3%, to less developmental damage than the injection of the patients found variant rs753661690 with 25.4% in the lower and 18.4% in the higher concentration. This cannot be found significant for both concentrations (p-value: > 0.05) (Fig. 4C).

Notably, taken together all heart phenotypes (combined morphological + cardiac and only cardiac phenotypes), the wt FAM222B mRNA led with a total amount of 12.9% of cardiac altered zf to less heart developmental damage than the injection of the patients found variant rs753661690 with 30.5% in the lower and 34.2% in the higher concentration (Fig. 4C and separately highlighted in 4C’). This can also be found significant for both concentrations (p-value: < 0.05*).

From these results, and especially underlining the fact that heart phenotypes occur significantly more often after our variants mRNA injection, we conclude that variant 1/2, rs753661690, p.300Arg > His found in our affected patient might very well lead to a disturbed cardiac development.