Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Mar 1.
Published in final edited form as: Birth Defects Res. 2024 Mar;116(3):e2327. doi: 10.1002/bdr2.2327

HUMAN SPLIT HAND/FOOT VARIANTS ARE NOT AS FUNCTIONAL AS WILDTYPE HUMAN PRDM1 IN THE RESCUE OF CRANIOFACIAL DEFECTS

Brittany T Truong 1,2, Lomeli C Shull 2,#, Bryan J Zepeda 4, Ezra Lencer 3, Kristin B Artinger 2,4
PMCID: PMC10949536  NIHMSID: NIHMS1972208  PMID: 38456586

Abstract

Background:

Split Hand/Foot Malformation (SHFM) is a congenital limb disorder presenting with limb anomalies, such as missing, hypoplastic, or fused digits, and often craniofacial defects, including a cleft lip/palate, microdontia, micrognathia, or maxillary hypoplasia. We previously identified three novel variants in the transcription factor, PRDM1, that are associated with SHFM phenotypes. One individual also presented with a high arch palate. Studies in vertebrates indicate that PRDM1 is important for development of the skull; however, prior to our study, human variants in PRDM1 had not been associated with craniofacial anomalies.

Methods:

Using transient mRNA overexpression assays in prdm1a−/− mutant zebrafish, we tested whether the PRDM1 SHFM variants were functional and could lead to a rescue of the craniofacial defects observed in prdm1a−/− mutants. We also mined previously published CUT&RUN and RNA-seq datasets that sorted EGFP-positive cells from a Tg(Mmu:Prx1-EGFP) transgenic line that labels the pectoral fin, pharyngeal arches, and dorsal part of the head to examine Prdm1a binding and the effect of Prdm1a loss on craniofacial genes.

Results:

prdm1a−/− mutants exhibit craniofacial defects including a hypoplastic neurocranium, a loss of posterior ceratobranchial arches, a shorter palatoquadrate, and an inverted ceratohyal. Injection of wildtype hPRDM1 in prdm1a−/− mutants partially rescues the palatoquadrate phenotype. However, injection of each of the three SHFM variants fails to rescue this skeletal defect. Loss of prdm1a leads to a decreased expression of important craniofacial genes by RNA-seq, including emilin3a, confirmed by HCR expression. Other genes including dlx5a/dlx6a, hand2, sox9b, col2a1a, and hoxb genes. Validation by qRT-PCR in the anterior half of zebrafish embryos failed to confirm the expression changes suggesting that the differences are enriched in prx1 expressing cells.

Conclusion:

These data suggest that the three SHFM variants are likely not functional and may be associated with the craniofacial defects observed in the humans. Finally, they demonstrate how Prdm1a can directly bind and regulate genes involved in craniofacial development.

Keywords: PRDM1, Split Hand/Foot Malformation, craniofacial, zebrafish

Introduction

Split Hand/Foot Malformation (SHFM) is a congenital limb disorder affecting 1 in 18,000 live births (Umair & Hayat, 2020). Individuals with SHFM present with missing, hypoplastic, and/or fused digits, though the phenotypes are highly variable due to incomplete penetrance. Many individuals also present with craniofacial defects, such as a cleft lip/palate, microdontia, micrognathia, and/or maxillary hypoplasia (reviewed in Sowinska-Seidler et al. (2014)). Pathogenic variants in WNT10B (MIM #225300), TP63 (MIM #605289), DLX5 (MIM #183600), ZAK (MIM #616890), EPS15L1 (MIM *616826), or chromosomal rearrangements in chromosomes 2 (MIM %606708), 10 (MIM #246560), or X (MIM %313350) are known to cause SHFM (Sowinska-Seidler et al., 2014). We recently identified three novel, pathogenic variants in a transcription factor, PRDM1, in individuals with SHFM: PRDM1c.712_713insT (p.C239Lfs*32); PRDM1c.1571C>G (p.T524R); and PRDM1c.2455A>G (p.T819A) (Truong et al., 2023). These variants negatively affect the protein’s ability to bind to DNA and regulate genes required for limb induction, outgrowth, differentiation, and anterior/posterior patterning (Truong et al., 2023). One individual also has a craniofacial defect. He came into the clinic with ectrodactyly ectodermal dysplasia (EEC) syndrome (MIM 129810) with bilateral 3/4-digit syndactyly and a high arch palate but no clefting. This individual has a de novo, missense mutation, PRDM1c.1571C>G (p.T524R).

Studies in zebrafish and mice indicate that PRDM1 is important for the development of the skull (Robertson et al., 2007; Vincent et al., 2005) (see below for details). Given the full null phenotype in mice, it is predicted that humans with null mutations would be lethal. Much of the craniofacial skeleton is derived from neural crest cells (NCCs), a multipotent population of cells derived from the non-neural ectoderm at the neural plate border. During development, these cells migrate away from the neural tube into different regions of the body and differentiate into various cell types, such as peripheral neurons, osteoblasts, melanocytes, and chondrocytes. Cranial NCCs migrate into the pharyngeal arches and facial prominences of the developing face. These cells will later contribute to the bone, cartilage, and connective tissue of the developing head skeleton.

In zebrafish, Prdm1a has been shown to be required for NCC specification and differentiation, as indicated by a significant decrease in NCC markers sox10, crestin, and snail2 in prdm1a−/− presumed null and hypomorph mutants (Artinger et al., 1999; Hernandez-Lagunas et al., 2005; Olesnicky et al., 2010; Roy & Ng, 2004). In turn, a loss of Prdm1a leads to a decrease in NCC derivatives, including melanocytes and cranial and dorsal root ganglia (Artinger et al., 1999; Birkholz et al., 2009; Hernandez-Lagunas et al., 2005; Olesnicky et al., 2010; Roy & Ng, 2004). prdm1a−/− mutants present with craniofacial defects, including an inverted ceratohyal, missing posterior ceratobranchial arches, and a shortened neurocranium (Birkholz et al., 2009). A hypoplastic neurocranium is indicative of orofacial clefting in zebrafish (reviewed in Truong and Artinger (2021)). In mice, Prdm1 is expressed in the endodermal layer of the first branchial arch at E9.5 and in the endoderm, ectoderm, and mesenchyme of the second and third arches (Robertson et al., 2007; Vincent et al., 2005). In both null Prdm1 mice and conditional Prdm1 knockouts in the embryo proper (Sox2:Cre), mutant embryos are able to properly form the first pharyngeal arch, and subsequently a lower jaw, but the more caudal arches are completely lost. Mutants are missing the thymus and exhibit hypoplasia of the pharyngeal epithelium (Robertson et al., 2007; Vincent et al., 2005). These studies provide evidence for the importance of PRDM1 in craniofacial development.

Prior to our study, human variants in PRDM1 had not been associated with craniofacial anomalies. Using transient overexpression assays, we sought to determine whether the PRDM1 SHFM variants were functional and could rescue the craniofacial defects in prdm1a−/− zebrafish.

Materials & Methods

Zebrafish husbandry

Zebrafish were maintained as previously described (Westerfield, 2000). The wildtype (WT) strain used was AB (ZIRC) and the mutant lines used were prdm1am805 (nrd; referred to as prdm1a−/−) (Artinger et al., 1999; Hernandez-Lagunas et al., 2005). All experiments were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Colorado Denver Anschutz Medical Campus (protocol #147) and conform to the NIH regulatory standards of care and treatment. Zebrafish lines can be obtained from the lead contact.

mRNA overexpression in zebrafish

mRNA overexpression was performed as previously described (Truong et al., 2023). hPRDM1 variant cDNA was synthesized into a pCS2+ backbone using Gateway cloning. cDNA was linearized and transcribed using the mMessage mMachine T7 Transcription Kit (ThermoFisher). prdm1a+/− fish were intercrossed and 10–15 pg of each hPRDM1 mRNA variants was injected into resulting embryos at the single-cell stage. Embryos were staged throughout the first four days of development to ensure there was no developmental delay or unassociated pathologies due to the mRNA overexpression. At 4 days post fertilization (dpf), larvae were collected for Alcian blue staining. Sample size refers to the number of individuals and is included in the figure legends.

Alcian blue cartilage staining

Zebrafish were stained for cartilage as previously described (Walker & Kimmel, 2007). In short, 4 dpf larvae were fixed in 2% paraformaldehyde (PFA) at room temperature for one hour. Larvae were then washed in 100mM Tris (pH 7.5)/10mM MgCl2 before rocking overnight at room temperature in Alcian blue stain (pH 7.5) (0.04% Alcian Blue, 80% ethanol, 100mM Tris [pH 7.5], 10mM MgCl2). Larvae were destained and rehydrated in a series of ethanol washes (80%, 50%, 25%) containing 100mM Tris (pH 7.5) and 10mM MgCl2, and then bleached for 10 minutes in 3% H2O2/0.5% KOH. Finally, larvae were rinsed twice in 25% glycerol/0.1% KOH to remove the bleach and stored at 4°C in 50% glycerol/0.1% KOH. The craniofacial skeletons of stained larvae were dissected, flat mounted in 50% glycerol/0.1 KOH, and imaged on an Olympus BX51 WI microscope. Measurements of the viscerocranium and neurocranium were performed blindly in ImageJ and then compared using a one-way ANOVA followed by a Tukey’s post-hoc test relative to uninjected prdm1a−/− mutants. Sample size refers to the number of individuals and is included in the figure legends.

Bioinformatics and RNA isolation for RT-qPCR

Our previous experiment of flow sorting, CUT&RUN, and RNA sequencing has been published (Truong et al., 2023). The bioinformatics analysis was performed as previously described therein. For a detailed description of the experimental design and analysis, we refer readers to Truong et al. 2023. Briefly, peak calling of CUT&RUN data was done using MACS2 using the default settings, and deepTools was used to generate the log fold change and p-values. The RNA-seq analysis was performed using STAR aligner and DESeq2. For RT-qPCR, total RNA was isolated from the heads of pooled wildtype and prdm1a−/− mutant embryos at 24 and 48 hours post fertilization (hpf) with TRIzol reagent (Invitrogen) and phenol/chloroform (5–10 embryo heads per biological replicate) as previously described (Truong et al., 2023). RNA (500 ng) was reverse transcribed to cDNA with SuperScript III First-Strand Synthesis System (Invitrogen) for real-time quantitative PCR (RT-qPCR). Primers for sox9b, col2a1a, emilin3a, dlx2a, and hand2 are: sox9b: (F) 5’-CAGCCCAGACGGAGGAAATC-3’, (R) 5’CCCTGAGACTGACCGGAGTG-3’; col2a1a: (F) 5’-GTGTGTGATTCGGGGACTGT-3’, (R) 5’-TTTGCACCAAGTGACCCGAT-3’; emilin3a: (F) 5’-TGCAGGAATCTCAGCATCACT-3’, (R) 5’-TATGCACAGTGGTTCTTGTGCC-3’; dlx2a: (F) 5’-ACAGTCTGCACAAGTCCCAG-3’, (R) 5’-TCGCTTTCTTCCTTTTCAGCAT-3’; hand2: (F) 5’-AGGCAGAAAGAAGGAAATGAATGAC-3’, (R) 5’-TCAGCTCCAATGCCCAAACA-3’; prdm1b: (F) 5’-GACTTTCTGTCACGAGCGGA-3’, (R) 5’-TAACCCACCCAACATCCTCC-3’; prdm1c: (F) 5’-CTCCTCCAGCACTACAAGCC-3’, (R) 5’-CAGAGAGCAGGACGATCCAC-3’.. Reactions were performed in at least three biological and technical replicates. Transcript abundance and relative fold change were quantified using the 2−ΔΔCt method relative to control. Relative expression was compared using an unpaired, independent t test. P-values less than 0.05 were considered statistically significant.

Hybridization Chain Reaction (HCR) V3.0

Probes for sox10 and emilin3a were purchased from Molecular Instruments (www.molecularinstruments.com). Whole mount HCR was performed according to the manufacturer’s instructions (Choi et al., 2016; Choi et al., 2018) and as previously described (Truong et al., 2023). Briefly, embryos were fixed overnight at 4°C in 4% PFA, washed in PBS, and dehydrated and permeabilized in two 10-minute washes in 100% MeOH at room temperature. Embryos were stored for at least 24 hours at −20°C in fresh MeOH. A graded series of MeOH/PBST solutions was used to rehydrate the embryos (75%, 50%, 25%, 0%). Embryos were then treated with proteinase K (10 μg/mL) for 5 minutes (24 hpf) or 15 minutes (48 hpf), washed twice in PBST, fixed for 20 minutes in 4% PFA, and then washed five times in PBST. Hairpins were heated to 95°C for 90 seconds and cooled. Embryos were stored in PBS at 4°C protected from light. Whole embryos were mounted in 0.2% low melt agarose and imaged on Zeiss LSM confocal. Embryos were then genotyped following Rossi et al. (2009) with slight modifications. Following DNA extraction, PCR was performed in M buffer (2mM MgCl2, 13.7 mM Tris-HCl [pH 8.4], 68.4mM KCl, 0.001% gelatin, 1.8 mg/mL protease-free BSA, 136 μM each dATP/CTP/GTP/TTP) with GoTaq Flexi (Promega) and digested overnight in Fok1 enzyme at 37°C.

Results

SHFM Human PRDM1 variants lead to craniofacial defects.

To determine whether the SHFM hPRDM1 variants are functional, we designed an in vivo rescue experiment. We overexpressed either wildtype hPRDM1 mRNA or each of the three SHFM variants in intercrossed prdm1a+/− zebrafish embryos and assessed whether there was a rescue to the craniofacial skeleton (Fig. 1). We have previously shown that overexpression of wildtype hPRDM1 mRNA rescues the pectoral fin of prdm1a−/− mutants (Truong et al., 2023), and injection of prdm1a zebrafish mRNA rescues NCC derivatives, including trunk NCC derivatives and Rohon-Beard sensory neurons (Hernandez-Lagunas et al., 2005). Uninjected prdm1a−/− mutants are missing posterior ceratobranchial arches, have an inverted and hypoplastic ceratohyal (16.28% decrease, p=0.0125), shorter palatoquadrate (21.57% decrease, p=0.0079), narrow ethmoid plate (14.46% decrease, p=0.0780), and a shortened neurocranium (16.56% decrease, p=0.0397) compared to uninjected wildtype/heterozygotes (Fig. 1A–B, H–K) (Birkholz et al., 2009). Injection of wildtype hPRDM1 partially rescues the length of the palatoquadrate (24.25% increase, p=0.0256) (Fig. 1C, I). Because this is the only cartilage element that is rescued, we focused on this element and found that injection of the three SHFM variants fails to rescue the palatoquadrate in this assay (Fig. 1C–F). The posterior ceratobranchial arches are consistently missing (Fig. 1C–G), and the ceratohyal, palatoquadrate, and ethmoid plate remained shorter (Fig. 1H–K). There is a slight increase in the length of the neurocranium from the ethmoid plate to the trabeculae upon injection with wildtype hPRDM1, PRDM1 (p.T524R), and PRDM1 (p.T819A), but the results were not statistically significant (Fig. 1J). Inability of the wildtype allele to fully rescue is likely due to the transience of the assay, insufficient dosing and/or rapid degradation of the mRNA. Together, these data suggest that the SHFM PRDM1 variants are not functional and may lead to craniofacial defects.

Figure 1: Transient overexpression of wildtype hPRDM1 in prdm1a−/− embryos rescues palatoquadrate, while the SHFM variants fail to rescue.

Figure 1:

Open in a new tab

prdm1a+/− heterozygous fish were intercrossed and injected with the hPRDM1 wildtype and SHFM variant mRNA at the single-cell stage. Injected larvae were collected at 4 dpf for Alcian blue staining. (A-F) Representative images of dissected craniofacial skeletons. The top row contains dissections of the viscerocranium, while the bottom row is the neurocranium. (A) Uninjected wildtype/heterozygous (n=10). (B) Uninjected prdm1a−/− mutant (n=6). prdm1a−/− mutants were injected with (C) wildtype hPRDM1 (n=10), (D) hPRDM1(p.C239Lfs*32) (n=10), (E) hPRDM1(p.T524R) (n=7), or (F) hPRDM1(p.T819A) mRNA (n=7). The asterisk (*) represents missing ceratobranchial arches. (G) Table showing the proportion of prdm1a−/− mutants that had missing posterior ceratobranchial arches #2–5. Measurements were taken to quantify the (H) length of the ceratohyal, (I) length of the palatoquadrate, (J) width of the ethmoid plate, and (K) length from the ethmoid plate to trabeculae. prdm1a−/− mutants have missing posterior ceratobranchial arches and a shorter ceratohyal, palatoquadrate, ethmoid plate, and neurocranium overall. Injection of wildtype hPRDM1 rescues the length of the palatoquadrate (p=0.0256), while the other variants do not. Injection of wildtype hPRDM1, hPRDM1(p.T524R), and hPRDM1(p.T819A) modestly rescues the length of the neurocranium. Abbreviations: cbs, ceratobranchial arches; ep, ethmoid plate; hs, hyosymplectic; m, Meckel’s cartilage; pq, palatoquadrate; tr, trabeculae

Loss of Prdm1a leads to decreased expression of craniofacial genes.

PRDM1 has been shown to be involved in vertebrate craniofacial development, though the mechanistic role remains unclear (Birkholz et al., 2009; Robertson et al., 2007; Vincent et al., 2005). We examined our previously published RNA-seq dataset from isolated EGFP-positive cells at 48 hours post fertilization (hpf) using a Tg(Mmu:Prx1-EGFP) transgenic line that labels the pectoral fin, pharyngeal arches, and dorsal part of the head (Truong et al., 2023). While we attempted to enrich for limb cells in the previous publication (Truong et al., 2023), it was clear that there were pharyngeal arch cells still present based on the transcripts that were differentially expressed. To determine if these transcripts were significantly differentially expressed, we replotted the data to show the fold change of craniofacial genes from the RNA-seq data (Fig. 2A).We found that genes known to be expressed in the pharyngeal arches or are involved in craniofacial development were downregulated, including sox10, hand2, dlx2a, dlx5a/dlx6a, gsc, barx1, prdm3, sox9b, col2a1a, and members of the hoxb gene family (Fig. 2A) (Dougherty et al., 2012; Miller et al., 2003; Robledo et al., 2002; Shull et al., 2020; Sperber et al., 2008; Yamada et al., 2021; Yan et al., 2005). Some of these transcripts are also expressed in the limb. One of the most significant downregulated genes in prdm1a−/− mutants was emilin3a, a glycoprotein within the extracellular matrix belonging to the EMILIN/multimerin family (Fig. 2A). This gene is expressed in the notochord, pharyngeal arches, and developing craniofacial skeleton of zebrafish, but was not identified to be expressed in the pectoral fin or limb in previous studies (Milanetto et al., 2007); however, its functional role in development has not yet been explored. Confirmation of expression by HCR in situ hybridization shows emilin3a is co-expressed with sox10 in the craniofacial region and is also highly expressed in the fin bud at 48 hpf. which is significantly reduced in prdm1a−/− mutants (Fig. 2B). This confirms that there is a decrease in craniofacial expression with the loss of prdm1a. Thus, it is possible that Prdm1a regulates expression of emilin3a or Prdm1a and Emilin3a interact during craniofacial and limb development. To validate other differentially expressed craniofacial genes from the RNA-seq data, we performed real-time quantitative PCR (RT-qPCR) on the anterior half of wildtype and prdm1a−/− mutant embryos at 24 and 48 hpf (Supp. Fig. 1). To examine any gene expression changes that may have occurred earlier before the RNA-seq was performed, we also assayed at 24 hpf. At this time point, only emilin3a was trending towards a significant decrease in prdm1a−/− mutants compared to wildtype (Supp. Fig. 1A). At 48 hpf, there is a trend towards a significant decrease in emilin3a, sox9b, and col2a1a expression (Supp. Fig. 1B). In addition, we assayed for genes that were upregulated in the RNA-seq, including prdm1a paralogs, prdm1b and prdm1c. Interestingly, only prdm1b is trending towards a significant increase at 48hpf suggesting that prdm1b may be compensating for the loss of prdm1a. These data suggest that in prx1+ cells, Prdm1a regulates gene expression, but this is not observed when assaying the anterior part of embryos.

Figure 2: RNA-seq results suggest loss of Prdm1a leads to a significant decrease in genes required for craniofacial development.

Figure 2:

Open in a new tab

RNA-seq was previously performed on Tg(Mmu:Prx1-EGFP) wildtype and prdm1a−/− embryos at 48 hpf (Truong et al. 2023). We reanalyzed the dataset and replotted the data. (A’) Bar plot showing the log2 fold change of differentially expressed genes in prdm1a−/− compared to wildtype. Orange bars are upregulated, while teal bars are downregulated. (A”) Bar plot showing the p-values of differentially expressed genes from RNA-seq dataset. (B) HCR in situ hybridization of emilin3a (green), sox10 (pink) and DAPI (blue) in wildtype and prdm1a−/− mutants at 48 hpf. White color indicates expression overlap. n=6 prdm1a mutants (C) CUT&RUN was previously performed on Tg(Mmu:Prx1-EGFP) wildtype embryos at 24 hpf (Truong et al. 2023). Tracks showing H3K27Ac enrichment (open chromatin) and Prdm1a binding sites for sox9b and col2a1a. Binding site sequence for Prdm1a, both the conserved core (GAAAG) and long sequence (AGYGAAAGYK), are shown as blue marks below the tracks. e, eye; f, fin; o, otic vesicle

We also examined the previously generated CUT&RUN dataset on the Tg(Mmu:Prx1-EGFP) transgenic line at 24 hpf to test direct Prdm1a binding to DNA. An earlier time point was used because Prdm1a must first bind to DNA before it can affect downstream gene expression. We mined the data for the known Prdm1 binding sites both a short conserved (GAAAG) and long sequence (AGYGAAAGYK) (Kuo & Calame, 2004) and found that Prdm1a likely directly binds putative enhancer and promoter regions within 1 kb upstream of sox9b, col2a1a, and dlx5a/dlx6a (Fig. 2C) (Truong et al., 2023). We did not, however, find any binding sites within 1 kb upstream of emilin3a. Given that these genes have decreased expression with a loss of prdm1a, it is possible that Prdm1a helps regulate cartilage formation in the craniofacial skeleton by directly binding these craniofacial targets. These data, together with previously published work, suggest that Prdm1a is required during craniofacial development.

Discussion

Although it is typically perceived as a congenital limb disorder, SHFM individuals have been reported to exhibit craniofacial anomalies, including cleft lip/palate, microdontia, micrognathia, and/or maxillary hypoplasia (reviewed in Sowinska-Seidler et al. (2014)). We recently identified three novel, causal SHFM variants in a gene PRDM1, which led to clefted hands/feet with variable penetrance (Truong et al., 2023). One individual also presented with a craniofacial defect, namely a high arched palate. We found that the PRDM1 variants have a dominant negative effect on the developing limb bud and fail to rescue the pectoral fin in prdm1a−/− mutant zebrafish, providing evidence for their pathogenicity (Truong et al., 2023). Here, we show that each of the three variants also fails to rescue the palatoquadrate of the craniofacial skeleton, while the overexpression of wildtype hPRDM1 mRNA partially restores the palatoquadrate, suggesting that the variants are not functional, at least in the morphogenesis of this structure. This was expected given that overexpression of zebrafish prdm1a mRNA sufficiently rescues trunk NCCs and Rohon-Beard sensory neurons in mutants (Hernandez-Lagunas et al., 2005). In that assay, although we did not specifically test for craniofacial cartilage, we predict that cranial NCCs would be rescued. The inability of the wildtype allele to rescue other elements of the craniofacial skeleton, such as the posterior ceratobranchial arches or hypoplastic neurocranium, may be because of the transient nature of the experiment or due to an insufficient dosage of PRDM1. We have previously shown that this dose is sufficient to rescue elements of the pectoral fin in prdm1a−/− mutants (Truong et al., 2023), suggesting that the assay is functional. However, it is interesting that the craniofacial defects are not as easily rescued. This may be due to the severity of the phenotype as well as its high penetrance (100%) (Fig. 1A, B, G). Our data supports the model that the SHFM variants are not functional in both the fin and the face. We also reanalyzed our previously published RNA-seq data comparing wildtype and prdm1a−/− mutants at 48 hpf and found a significant decrease in genes critical for craniofacial development, including dlx5a/dlx6a, hand2, barx1, gsc, sox9b, col2a1a, and members of the hoxb gene family (Fig. 2A) (Truong et al., 2023). We attempted to validate these results using RT-qPCR; however, because we used head cells instead of sorted cells, they are not significant (Supp. Fig. 1). Our analysis of previously published CUT&RUN data at 24 hpf also suggests that Prdm1a is likely directly binding to promoters and putative enhancers of sox9b, and col2a1a (Fig. 2C) (Truong et al., 2023).

The experiments presented here have strengths and weaknesses. First, we are using a transient level of RNA to rescue a zebrafish mutant. The strength of this assay is that we can rapidly determine if a given transcript is able to rescue a phenotype. This allows us to test the functionality of gene domains or human variants, as done here. In contrast, while this type of assay is used in many developmental systems, the transient and potentially mosaic nature of the experiment can be challenging to interpret. Nevertheless, this is a powerful assay to test functionality. Second, we are drawing conclusions from a previously published RNA-seq dataset in which we attempted to enrich for pectoral fin cells using prx1 expression. This is not optimized for NCCs, and thus, not the best way to conduct this experiment. In the future, we will repeat these assays in sorted NCCs so that stronger conclusions can be made. It is interesting that we were still able to identify specific craniofacial targets in our analysis, and these can be studied in the future. Third, while validating the RNA-seq targets, we used anterior portions of the zebrafish embryo. Because this is not the same as the FAC-sorted population, it is difficult to compare these results to the RNA-seq directly. However, together with the RNA-seq, we feel that this supports the hypothesis that there are differences in prdm1a function between the prx1+ cells and the rest of the embryo.

Of the differentially expressed craniofacial genes identified in prdm1a mutants, there are several genes of particular interest. One of the most significant downregulated genes in prdm1a−/− mutants was emilin3a, a glycoprotein within the extracellular matrix belonging to the EMILIN/multimerin family. This gene is expressed in the notochord, pharyngeal arches, and developing craniofacial skeleton of zebrafish (Corallo et al., 2013; Milanetto et al., 2007); however, its role has not yet been explored. emilin3a expression is reduced in prdm1a−/− embryos, but there are no Prdm1a binding sites in the 1 kb region we focused on. Therefore, it is possible that there are binding sites outside of this region; Prdm1a may indirectly regulate emilin3a expression; or that Prdm1a and Emilin3a interact indirectly during craniofacial development. Another interesting gene is dlx5a/dlx6a, a set of homeotic genes required for both limb development and dorsal/ventral patterning in the face (Robledo et al., 2002). Variants in DLX5 are known to cause SHFM Type I (MIM #183600) with orofacial clefting (Bernardini et al., 2008; Elliott & Evans, 2006), and we have shown that Prdm1a directly binds to and regulates dlx5a in the pectoral fin (Truong et al., 2023). We hypothesize that Prdm1a is required for maintaining proper patterning in the face through its regulation of dlx5a/dlx6a in the pharyngeal arches as well, though additional experiments are needed to determine this. Structures in the craniofacial and limb skeleton are clearly distinct from one another, but they utilize many of the same gene regulatory networks and mechanisms for their development (Truong & Artinger, 2021). This then leads to a frequent co-occurrence of craniofacial and limb anomalies in congenital diseases. These future studies will provide critical insight into the mechanism by which Prdm1a regulates both craniofacial and limb development and how disruptions to the gene regulatory networks involved can lead to SHFM.

Supplementary Material

Fig S1

Supplemental Figure 1: Loss of Prdm1a leads to a decrease in gene expression. (A-B) RT-qPCR was performed in pooled embryo heads at (A) 24 hpf and (B) 48 hpf in wildtype and prdm1a−/− mutants for dlx2a, hand2, emilin3a, sox9b, col2a1a, prdm1b, and prdm1c (n=5–6 embryos per genotype per biological replicate. Three biological replicates were used). Relative expression was compared using an unpaired, independent Student’s t-test. Error bars represent the mean ± SD. At 24 hpf, there is a slight decrease in emilin3a expression in prdm1a−/− mutants. At 48 hpf, there is a slight decrease in emilin3a, sox9b, and col2a1a and an increase in prdm1b in prdm1a−/− mutants.

Acknowledgements:

We thank members of the Artinger Lab for project feedback; Christine Archer and the zebrafish facility team for excellent animal care; and the SHFM families for participation in the study.

Funding statement:

This work is supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) (1F31HD103368 to B.T.T. and 1R03HD096320-01A1 to K.B.A.) and, in part, by the South Carolina Department of Disabilities and Special Needs. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Footnotes

Conflict of interest disclosure: We have no conflicts to declare.

Data Availability:

The RNA-seq and CUT&RUN data have been deposited in NCBI’s Gene Expression Omnibus and are accessible through accession number GSE217486 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE217486).

REFERENCES

  1. Artinger KB, Chitnis AB, Mercola M, & Driever W (1999). Zebrafish narrowminded suggests a genetic link between formation of neural crest and primary sensory neurons. Development, 126(18), 3969–3979. http://www.ncbi.nlm.nih.gov/pubmed/10457007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bernardini L, Palka C, Ceccarini C, Capalbo A, Bottillo I, Mingarelli R, Novelli A, & Dallapiccola B (2008). Complex rearrangement of chromosomes 7q21.13-q22.1 confirms the ectrodactyly-deafness locus and suggests new candidate genes. American Journal of Medical Genetics Part A, 146A(2), 238–244. https://doi.org/ 10.1002/ajmg.a.32093 [DOI] [PubMed] [Google Scholar]
  3. Birkholz DA, Olesnicky Killian EC, George KM, & Artinger KB (2009). Prdm1a is necessary for posterior pharyngeal arch development in zebrafish. Dev Dyn, 238(10), 2575–2587. 10.1002/dvdy.22090 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Choi HM, Calvert CR, Husain N, Huss D, Barsi JC, Deverman BE, Hunter RC, Kato M, Lee SM, Abelin AC, Rosenthal AZ, Akbari OS, Li Y, Hay BA, Sternberg PW, Patterson PH, Davidson EH, Mazmanian SK, Prober DA, … Pierce NA (2016). Mapping a multiplexed zoo of mRNA expression. Development, 143(19), 3632–3637. 10.1242/dev.140137 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Choi HMT, Schwarzkopf M, Fornace ME, Acharya A, Artavanis G, Stegmaier J, Cunha A, & Pierce NA (2018). Third-generation in situ hybridization chain reaction: multiplexed, quantitative, sensitive, versatile, robust. Development, 145(12). 10.1242/dev.165753 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Corallo D, Schiavinato A, Trapani V, Moro E, Argenton F, & Bonaldo P (2013). Emilin3 is required for notochord sheath integrity and interacts with Scube2 to regulate notochord-derived Hedgehog signals. Development, 140(22), 4594–4601. https://doi.org/ 10.1242/dev.094078 [DOI] [PubMed] [Google Scholar]
  7. Dougherty M, Kamel G, Shubinets V, Hickey G, Grimaldi M, & Liao EC (2012). Embryonic Fate Map of First Pharyngeal Arch Structures in the sox10: kaede Zebrafish Transgenic Model. Journal of Craniofacial Surgery, 23(5), 1333–1337. 10.1097/SCS.0b013e318260f20b [DOI] [PubMed] [Google Scholar]
  8. Elliott AM, & Evans JA (2006). Genotype–phenotype correlations in mapped split hand foot malformation (SHFM) patients. American Journal of Medical Genetics Part A, 140A(13), 1419–1427. https://doi.org/ 10.1002/ajmg.a.31244 [DOI] [PubMed] [Google Scholar]
  9. Hernandez-Lagunas L, Choi IF, Kaji T, Simpson P, Hershey C, Zhou Y, Zon L, Mercola M, & Artinger KB (2005). Zebrafish narrowminded disrupts the transcription factor prdm1 and is required for neural crest and sensory neuron specification. Dev Biol, 278(2), 347–357. 10.1016/j.ydbio.2004.11.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Kuo TC, & Calame KL (2004). B Lymphocyte-Induced Maturation Protein (Blimp)-1, IFN Regulatory Factor (IRF)-1, and IRF-2 Can Bind to the Same Regulatory Sites1. The Journal of Immunology, 173(9), 5556–5563. 10.4049/jimmunol.173.9.5556 [DOI] [PubMed] [Google Scholar]
  11. Milanetto M, Tiso N, Braghetta P, Volpin D, Argenton F, & Bonaldo P (2007). Emilin genes are duplicated and dynamically expressed during zebrafish embryonic development. Dev Dyn, 237(1), 222–232. https://doi.org/ 10.1002/dvdy.21402 [DOI] [PubMed] [Google Scholar]
  12. Miller CT, Yelon D, Stainier DYR, & Kimmel CB (2003). Two endothelin 1 effectors, hand2 and bapx1, pattern ventral pharyngeal cartilage and the jaw joint. Development, 130, 1353–1365. 10.1242/dev.00339 [DOI] [PubMed] [Google Scholar]
  13. Olesnicky E, Hernandez-Lagunas L, & Artinger KB (2010). prdm1a Regulates sox10 and islet1 in the development of neural crest and Rohon-Beard sensory neurons. Genesis, 48(11), 656–666. 10.1002/dvg.20673 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Robertson EJ, Charatsi I, Joyner CJ, Koonce CH, Morgan M, Islam A, Paterson C, Lejsek E, Arnold SJ, Kallies A, Nutt SL, & Bikoff EK (2007). Blimp1 regulates development of the posterior forelimb, caudal pharyngeal arches, heart and sensory vibrissae in mice. Development, 134(24), 4335–4345. 10.1242/dev.012047 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Robledo RF, Rajan L, Li X, & Lufkin T (2002). The Dlx5 and Dlx6 homeobox genes are essential for craniofacial, axial, and appendicular skeletal development. Genes Dev, 16(9), 1089–1101. 10.1101/gad.988402 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Rossi CC, Kaji T, & Artinger KB (2009). Transcriptional control of Rohon-Beard sensory neuron development at the neural plate border. Dev Dyn, 238(4), 931–943. 10.1002/dvdy.21915 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Roy S, & Ng T (2004). Blimp-1 specifies neural crest and sensory neuron progenitors in the zebrafish embryo. Curr Biol, 14, 1772–1777. 10.1016/j.cub.2004.09.046 (1772) [DOI] [PubMed] [Google Scholar]
  18. Shull LC, Sen R, Menzel J, Goyama S, Kurokawa M, & Artinger KB (2020). The conserved and divergent roles of Prdm3 and Prdm16 in zebrafish and mouse craniofacial development. Dev Biol, 461(2), 132–144. 10.1016/j.ydbio.2020.02.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Sowinska-Seidler A, Socha M, & Jamsheer A (2014). Split-hand/foot malformation - molecular cause and implications in genetic counseling. J Appl Genet, 55(1), 105–115. 10.1007/s13353-013-0178-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Sperber SM, Saxena V, Hatch G, & Ekker M (2008). Zebrafish dlx2a contributes to hindbrain neural crest survival, is necessary for differentiation of sensory ganglia and functions with dlx1a in maturation of the arch cartilage elements. Developmental Biology, 314(1), 59–70. https://doi.org/ 10.1016/j.ydbio.2007.11.005 [DOI] [PubMed] [Google Scholar]
  21. Truong BT, & Artinger KB (2021). The power of zebrafish models for understanding the co-occurrence of craniofacial and limb disorders. Genesis. 10.1002/dvg.23407 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Truong BT, Shull LC, Lencer E, Bend EG, Field M, Blue EE, Bamshad MJ, Skinner C, Everman D, Schwartz CE, Flanagan-Steet H, & Artinger KB (2023). PRDM1 DNA-binding zinc finger domain is required for normal limb development and is disrupted in split hand/foot malformation. Disease Models & Mechanisms, 16(4). 10.1242/dmm.049977 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Umair M, & Hayat A (2020). Nonsyndromic Split-Hand/Foot Malformation: Recent Classification. Mol Syndromol, 10(5), 243–254. 10.1159/000502784 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Vincent SD, Dunn NR, Sciammas R, Shapiro-Shalef M, Davis MM, Calame K, Bikoff EK, & Robertson EJ (2005). The zinc finger transcriptional repressor Blimp1/Prdm1 is dispensable for early axis formation but is required for specification of primordial germ cells in the mouse. Development, 132(6), 1315–1325. 10.1242/dev.01711 [DOI] [PubMed] [Google Scholar]
  25. Walker M, & Kimmel C (2007). A two-color acid-free cartilage and bone stain for zebrafish larvae. Biotechnic & Histochemistry, 82(1), 23–28. https://doi.org/ 10.1080/10520290701333558 [DOI] [PubMed] [Google Scholar]
  26. Westerfield M (2000). The zebrafish book. A guide for the laboratory use of zebrafish (Danio rerio) (4 ed.) [Online Book]. University of Oregon Press. https://zfin.org/zf_info/zfbook/zfbk.html [Google Scholar]
  27. Yamada K, Maeno A, Araki S, Kikuchi M, Suzuki M, Ishizaka M, Satoh K, Akama K, Kawabe Y, Suzuki K, Kobayashi D, Hamano N, & Kawamura A (2021). An atlas of seven zebrafish hox cluster mutants provides insights into sub/neofunctionalization of vertebrate Hox clusters. Development, 148(11). 10.1242/dev.198325 [DOI] [PubMed] [Google Scholar]
  28. Yan Y-L, Willoughby J, Liu D, Crump JG, Wilson C, Miller CT, Singer A, Kimmel C, Westerfield M, & Postlethwait JH (2005). A pair of Sox: distinct and overlapping functions of zebrafish sox9 co-orthologs in craniofacial and pectoral fin development. Development, 132(5), 1069–1083. 10.1242/dev.01674 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Fig S1

Supplemental Figure 1: Loss of Prdm1a leads to a decrease in gene expression. (A-B) RT-qPCR was performed in pooled embryo heads at (A) 24 hpf and (B) 48 hpf in wildtype and prdm1a−/− mutants for dlx2a, hand2, emilin3a, sox9b, col2a1a, prdm1b, and prdm1c (n=5–6 embryos per genotype per biological replicate. Three biological replicates were used). Relative expression was compared using an unpaired, independent Student’s t-test. Error bars represent the mean ± SD. At 24 hpf, there is a slight decrease in emilin3a expression in prdm1a−/− mutants. At 48 hpf, there is a slight decrease in emilin3a, sox9b, and col2a1a and an increase in prdm1b in prdm1a−/− mutants.

NIHMS1972208-supplement-Fig_S1.png (886.2KB, png)

Data Availability Statement

The RNA-seq and CUT&RUN data have been deposited in NCBI’s Gene Expression Omnibus and are accessible through accession number GSE217486 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE217486).

RESOURCES