Connexin 43 Gene Ablation Does Not Alter Human Pluripotent Stem Cell Germ Lineage Specification

During embryonic germ layer development, cells communicate with each other and their environment to ensure proper lineage specification and tissue development. Connexin (Cx) proteins facilitate direct cell–cell communication through gap junction channels. While previous reports suggest that gap junctional intercellular communication may contribute to germ layer formation, there have been limited comprehensive expression analyses or genetic ablation studies on Cxs during human pluripotent stem cell (PSC) germ lineage specification. We screened the mRNA profile and protein expression patterns of select human Cx isoforms in undifferentiated human induced pluripotent stem cells (iPSCs), and after directed differentiation into the three embryonic germ lineages: ectoderm, definitive endoderm, and mesoderm. Transcript analyses by qPCR revealed upregulation of Cx45 and Cx62 in iPSC-derived ectoderm; Cx45 in mesoderm; and Cx30.3, Cx31, Cx32, Cx36, Cx37, and Cx40 in endoderm relative to control human iPSCs. Generated Cx43 (GJA1) CRISPR-Cas9 knockout iPSCs successfully differentiated into cells of all three germ layers, suggesting that Cx43 is dispensable during directed iPSC lineage specification. Furthermore, qPCR screening of select Cx transcripts in our GJA1-/- iPSCs showed no significant Cx upregulation in response to the loss of Cx43 protein. Future studies will reveal possible compensation by additional Cxs, suggesting targets for future CRISPR-Cas9 ablation studies in human iPSC lineage specification.


Stem Cell Differentiation
Directed iPSC differentiation toward definitive endoderm, mesoderm, or ectoderm was performed using the STEMdiff™ Trilineage Differentiation Kit (Cat# 05230, STEMCELL Technologies), according to the manufacturer's instructions.
For spontaneous monolayer differentiation, iPSCs were seeded into Geltrex-coated cell culture vessels containing Essential 6 media (Cat# A1516401, ThermoFisher), which lacks the crucial pluripotency growth factors TGF-β and FGF2 [31]. Media were changed every 1-2 days during the spontaneous differentiation process [32], and differentiation was allowed to progress until day nine before processing.
For spontaneous differentiation via embryoid bodies (EBs), 9000 Accutase-digested single-celled iPSCs in Essential 6 fortified with 10 µM of Y-27632 were plated into a 96well round-bottom plate coated with 1% agarose prepared in deionized water to confer a non-adherent surface [33,34]. EBs were fed with Essential 6 every other day to promote spontaneous differentiation.

Confocal Microscopy
Confocal immunofluorescence images were acquired on an Olympus Fluoview FV10i-W3 confocal microscope (Olympus, Tokyo, Japan) running Fluoview v2.1.1.7 software, equipped with 60X/1.2 NA and 10X/0.4 NA water immersion lenses, and the following lasers to visualize fluorophores: DAPI/Hoechst 33342 (405-nm laser); Alexa Fluor 488/eGFP (473-nm laser); Phalloidin/Alexa Fluor 555 (559-nm laser); and Alexa Fluor 647 (635-nm laser). Laser power and sensitivity were adjusted to visualize immunofluorescence and minimize background signal. Additional sample imaging was performed on an Olympus Fluoview FV1000 confocal microscope fitted with 10X/0.4 NA, 20X/0.75 NA or 40X/0.95 NA lenses and the following lasers: 405 nm, 458 nm, 568 nm, and 633 nm. Images were analyzed using Fiji open-source software. Images were analyzed, pseudo-colored, and made into composites with Fiji software [36]. Brightness and contrast were equally adjusted for optimal visualization in all images using Fiji software. Quantitative analysis of protein colocalization was performed with the JACoP plugin for Fiji which reports Manders' coefficient values, as described in Bolte et al. (2006) [37].

RNA Transcript Analysis
RNA extraction was conducted using the PureLink™ RNA Mini Kit (Cat# 12183025, Thermofisher). Five hundred nanograms of extracted RNA was converted to complementary DNA (cDNA) using the High-Capacity cDNA Reverse Transcription Kit (Cat# 4368814, Thermofisher). Quantitative reverse transcription-polymerase chain reaction (qPCR) analysis was performed using intercalating dye technology (ssoAdvanced SYBR green supermix) (Cat# 1725274, BioRad), according to the manufacturer's instructions and an annealing temperature of 60 • C. Primer sets were purchased from Integrated DNA Technologies (Newark, NJ, USA) where catalog information and sequences are listed in Table 3. PCR reactions were run on a ViiA™ 7 (ThermoFisher) running QuantStudio Real-Time PCR software version 1.3. Table 3. Primer sets used in qPCR analyses.

Target
Forward
Fluorescent images of dye transfer in iPSCs, ectoderm, mesoderm, or endoderm were taken using a Zeiss AxioObserver microscope equipped with an X-Cite Series 120Q lamp (Excelitas Technologies, Mississauga, ON, Canada). Image acquisition used the 5X/0.12 NA A-Plan and 10X/0.25 NA Ph1 objectives. Pseudo-coloring, scale calibration, and analysis of dye transfer was performed with Fiji software [36]. Four or more images were taken per sample, and ten dye migration measurements made per image. Brightness and contrast were equally adjusted for optimal visualization in all images using Fiji software.

Flow Cytometry
Flow cytometry of singularized cells from embryoid bodies was performed on a CytoFLEX (Beckman Coulter, Brea, CA, USA) flow cytometer. Antibodies were titrated over a range of concentrations prior to use and the following controls were included in all flow cytometry assays: unstained control, fluorescence-minus-one (FMO) controls, and single-color compensation controls for fluorochromes. UltraComp compensation beads (Cat# 01-2222-43, ThermoFisher) were used with antibodies raised in mice.
Live single-cell suspensions were labelled with Zombie NIR™ fixable viability dye (Cat# 423105, BioLegend ® , San Diego, CA, USA) diluted at 1:1000 to eliminate dead cells during the analysis stage. Next, the cells were fixed in 10% normal buffered formalin for 10 minutes at 4 • C followed by permeabilization for 15 minutes at room temperature with Ca 2+ and Mg 2+ -free PBS supplemented with 0.5% BSA and 0.1% Triton X-100. Primary antibodies (used at dilutions according to Table 4) were incubated for 30 min at 4 • C in the dark. Flow cytometric analysis was performed using FlowJo software (version 10.7.1).

Statistical Analysis
Statistical analysis and plotting of raw data were performed using Graph Pad Prism v.8. Graphs presented as ± standard error of the mean (SEM). Unless otherwise stated, n ≥ 3 independent biological replicates. Student's t-test was performed for statistical analysis between two groups. Larger data sets of three or more groups were analyzed using analysis of variance (ANOVA) with Tukey's multiple comparisons test where * p < 0.05, ** p < 0.01, and *** p < 0.001.

Cx43 Protein Expression Persists throughout Lineage Specification
In line with the broad expression of Cx43 throughout the developing embryo, we found comparable Cx43 transcript expression across control iPSCs and all three embryonic germ lineages (Figure 1). Due to the widely reported role of Cx43 in stem cell differentiation [12,20,22,27,29], along with our findings that Cx43 transcripts are similarly expressed during iPSC lineage commitment, we focused on Cx43 protein expression and function across cells of the three germ lineages. Confocal microscopy and Western blot analyses of lineage-specific proteins demonstrate the successful differentiation of iPSCs to ectoderm (PAX6 or Nestin), endoderm (SOX17), and mesoderm (Brachyury) (Figure 2). Additionally, immunofluorescence confocal microscopy revealed Cx43 localized as small puncta, indicative of gap junction plaques, at the cellular interfaces of iPSCs and of each embryonic germ lineage (Figure 2A). Total Cx43 protein was not significantly different across undifferentiated iPSCs, ectodermal, or endodermal cells. However, in contrast to our qPCR results, iPSC differentiation toward mesoderm cells was accompanied by a significant increase in Cx43 protein compared to undifferentiated iPSCs ( Figure 2B,C).

Human iPSCs Tolerate GJA1 CRISPR-Cas9 Editing
Previous studies demonstrated that Cx43 was dispensable for human iPSC survival and pluripotency gene expression [20]. However, due to the wide expression profile of

Human iPSCs Tolerate GJA1 CRISPR-Cas9 Editing
Previous studies demonstrated that Cx43 was dispensable for human iPSC survival and pluripotency gene expression [20]. However, due to the wide expression profile of Cx43 throughout early development, we sought to determine whether this protein plays a role in early cell fate specification. In addition to our previously published GJA1-/-(Cx43 knockout) iPSC line [20,28], we utilized Cx43-eGFP reporter iPSCs generated by the Allen Institute for Cell Science. This reporter line harbors a heterozygous insertion of the Cx43-eGFP construct at the endogenous GJA1 allele, resulting in iPSCs with one wildtype GJA1 allele and one genetically altered allele expressing Cx43-eGFP ( Figure 3A). Importantly, the Cx43-eGFP iPSCs do not overexpress the reporter construct as the inserted Cx43-eGFP construct remains under control of the endogenous Cx43 promoter. Western blotting revealed bands of appropriate size for Cx43 (~43 kDa) in control iPSCs, which is absent in GJA1-/-iPSCs ( Figure 3A). On the other hand, the Cx43-eGFP reporter cell line exhibits two distinct protein species corresponding to the endogenous Cx43 allele (~43 kDa) and the Cx43-eGFP knock-in allele (~60 kDa) ( Figure 3A). Immunofluorescent confocal microscopy confirmed Cx43 protein expression at the cell surface in control human iPSCs, where Cx43 forms large puncta at opposing cell membranes ( Figure 3B). As expected, punctate staining patterns are absent in GJA1-/-iPSCs ( Figure 3B). Our Cx43-eGFP reporter iPSCs display comparable Cx43 expression and localization to control iPSCs, exhibiting large plasma membrane puncta ( Figure 3B). Furthermore, we noted no observable difference in cellular morphology of the edited iPSCs compared to control ( Figure 3B). Taken together, these three human iPSC lines (control, GJA1-/-, and Cx43-eGFP) enable a comprehensive evaluation of Cx43 during iPSC lineage commitment and differentiation.

Cx43-eGFP iPSCs Differentiate into All Three Germ Lineages
While the Cx43-eGFP human iPSC line exhibits comparable Cx43 expression levels and subcellular localization to control cells, it remained important to confirm that these genetically engineered iPSCs retained the ability to differentiate to the three embryonic germ layers. To that end, we performed 2-dimensional spontaneous monolayer iPSC differentiation to compare the inherent differentiation potential of the control iPSCs ( Figure 4A) and Cx43-eGFP iPSCs ( Figure 4B). Unlike directed differentiation, spontaneous differentiation does not utilize exogenous signals to drive cells toward specific lineages but instead relies on intrinsic cellular communication to confer fate decisions. Like control iPSCs, spontaneously differentiated Cx43-eGFP iPSCs successfully produced cells of each embryonic germ lineage, as identified by Nestin (ectoderm), Brachyury (mesoderm), and SOX17 (endoderm) expression ( Figure 4A,B). Additionally, both control and Cx43-eGFPdifferentiated cells expressed Cx43 across cells of each germ layer and formed the typical puncta indicative of gap junction plaques ( Figure 4A,B). display comparable Cx43 expression and localization to control iPSCs, exhibiting large plasma membrane puncta ( Figure 3B). Furthermore, we noted no observable difference in cellular morphology of the edited iPSCs compared to control ( Figure 3B). Taken together, these three human iPSC lines (control, GJA1-/-, and Cx43-eGFP) enable a comprehensive evaluation of Cx43 during iPSC lineage commitment and differentiation.  instead relies on intrinsic cellular communication to confer fate decisions. Like control iPSCs, spontaneously differentiated Cx43-eGFP iPSCs successfully produced cells of each embryonic germ lineage, as identified by Nestin (ectoderm), Brachyury (mesoderm), and SOX17 (endoderm) expression ( Figure 4A,B). Additionally, both control and Cx43-eGFPdifferentiated cells expressed Cx43 across cells of each germ layer and formed the typical puncta indicative of gap junction plaques ( Figure 4A,B).

Cx43 Is Dispensable during Lineage Specification
Previous reports using pharmacological gap junction blockers or Cx43 siRNA knockdown suggest that Cx43 influences human and mouse PSC germ lineage specification [22,27]. To determine whether our GJA1-/-iPSCs exhibit similar germ lineage biases, we directed both control and GJA1-/-iPSCs to form ectoderm, endoderm, or mesoderm using a commercially available kit ( Figure 5). In contrast to previous studies, we found no significant difference in ectoderm (PAX6; PAX6, NES), mesoderm (Brachyury; T, MIXL1, NCAM1), or endoderm (SOX17; SOX17, FOXA2) cells differentiated from either control or GJA1-/-iPSCs ( Figure 5). In addition to directed differentiation, we compared the inherent differentiation preferences of control and GJA1-/-iPSCs using a 3-dimensional embryoid body (EB) model of spontaneous differentiation ( Figure 6). Control and GJA1-/-iPSCs both successfully self-aggregated and formed EBs of comparable size when cultured in nonadherent cell culture dishes ( Figure 6A). After 14 days of spontaneous differentiation, qPCR and flow cytometry revealed no significant difference in germ lineage-specific

Cx43 Is Dispensable during Lineage Specification
Previous reports using pharmacological gap junction blockers or Cx43 siRNA knockdown suggest that Cx43 influences human and mouse PSC germ lineage specification [22,27]. To determine whether our GJA1-/-iPSCs exhibit similar germ lineage biases, we directed both control and GJA1-/-iPSCs to form ectoderm, endoderm, or mesoderm using a commercially available kit ( Figure 5). In contrast to previous studies, we found no significant difference in ectoderm (PAX6; PAX6, NES), mesoderm (Brachyury; T, MIXL1, NCAM1), or endoderm (SOX17; SOX17, FOXA2) cells differentiated from either control or GJA1-/-iPSCs ( Figure 5). In addition to directed differentiation, we compared the inherent differentiation preferences of control and GJA1-/-iPSCs using a 3-dimensional embryoid body (EB) model of spontaneous differentiation ( Figure 6). Control and GJA1-/-iPSCs both successfully self-aggregated and formed EBs of comparable size when cultured in non-adherent cell culture dishes ( Figure 6A). After 14 days of spontaneous differentiation, qPCR and flow cytometry revealed no significant difference in germ lineage-specific marker expression in EBs formed using control or GJA1-/-iPSCs ( Figure 6B,C). Therefore, as GJA1-/-iPSCs successfully differentiated into all three germ lineages under both directed and spontaneous conditions, we conclude that Cx43 does not appear to skew human iPSC germ lineage specification. Inclusion of additional trophectoderm-containing models, such as gastruloids or blastoids, would allow future studies to further investigate the influence of Cxs on early cell signaling events regulating subsequent organization and patterning.
as GJA1-/-iPSCs successfully differentiated into all three germ lineages under both directed and spontaneous conditions, we conclude that Cx43 does not appear to skew human iPSC germ lineage specification. Inclusion of additional trophectoderm-containing models, such as gastruloids or blastoids, would allow future studies to further investigate the influence of Cxs on early cell signaling events regulating subsequent organization and patterning.

Other Connexin Isoforms Do Not Compensate for the Loss of Cx43 during Lineage Specification
As described above, many connexin isoforms are expressed in human iPSCs and several are dynamically regulated during iPSC fate specification. Given that iPSCs, ectoderm, mesoderm, and endoderm lineages all tolerate the loss of Cx43 without apparent detriment ( Figure 6), we investigated whether other Cx isoforms are upregulated to compensate for GJA1 ablation. We found no significant difference in Cx transcript expression between control and GJA1-/-cells in any of the lineages investigated (Figure 7). Using gap

Other Connexin Isoforms Do Not Compensate for the Loss of Cx43 during Lineage Specification
As described above, many connexin isoforms are expressed in human iPSCs and several are dynamically regulated during iPSC fate specification. Given that iPSCs, ectoderm, mesoderm, and endoderm lineages all tolerate the loss of Cx43 without apparent detriment (Figure 6), we investigated whether other Cx isoforms are upregulated to compensate for GJA1 ablation. We found no significant difference in Cx transcript expression between control and GJA1-/cells in any of the lineages investigated (Figure 7). Using gap junction-permeable dyes, Lucifer yellow ( Figure 8A), and neurobiotin ( Figure 8B), we performed scrape loading dye transfer assays to assess gap junction function in control and GJA1-/-iPSCs as well as the three germ layers. Control iPSCs exhibited an average Lucifer yellow dye transfer distance of 83.33 ± 6.14 µm and 221.1 ± 30.80 µm for the smaller neurobiotin tracer. In control ectoderm, we observed dye transfer distances of 50.47 ± 1.77 µm for Lucifer yellow and 100.90 ± 21.02 µm for neurobiotin. Meanwhile, control mesoderm cells displayed average dye transfer distances of 70.95 ± 11.59 µm for Lucifer yellow and 145.70 ± 7.63 µm for neurobiotin, and, finally, control endoderm-differentiated cells displayed 89.03 ± 16.64 µm Lucifer yellow transfer and 193.40 ± 46.31 µm for neurobiotin. In contrast to control cells, GJA1-/-iPSCs cells exhibited significantly reduced dye transfer in undifferentiated iPSCs, and after differentiation into each germ lineage relative to control counterparts. Together, these qPCR and dye transfer analyses suggest that GJA1-/-iPSCs do not upregulate other Cx isoforms to compensate for the loss of Cx43, even when directed to differentiate toward the three embryonic germ lineages.
Biomolecules 2021, 11, x 14 of 21 junction-permeable dyes, Lucifer yellow ( Figure 8A), and neurobiotin ( Figure 8B), we performed scrape loading dye transfer assays to assess gap junction function in control and GJA1-/-iPSCs as well as the three germ layers. Control iPSCs exhibited an average Lucifer yellow dye transfer distance of 83.33 ± 6.14 µ m and 221.1 ± 30.80 µ m for the smaller neurobiotin tracer. In control ectoderm, we observed dye transfer distances of 50.47 ± 1.77 µ m for Lucifer yellow and 100.90 ± 21.02 µ m for neurobiotin. Meanwhile, control mesoderm cells displayed average dye transfer distances of 70.95 ± 11.59 µ m for Lucifer yellow and 145.70 ± 7.63 µ m for neurobiotin, and, finally, control endoderm-differentiated cells displayed 89.03 ± 16.64 µ m Lucifer yellow transfer and 193.40 ± 46.31 µ m for neurobiotin. In contrast to control cells, GJA1-/-iPSCs cells exhibited significantly reduced dye transfer in undifferentiated iPSCs, and after differentiation into each germ lineage relative to control counterparts. Together, these qPCR and dye transfer analyses suggest that GJA1-/-iPSCs do not upregulate other Cx isoforms to compensate for the loss of Cx43, even when directed to differentiate toward the three embryonic germ lineages.
Like previous reports, we find that Cx43 is expressed in undifferentiated iPSCs and can be readily detected at the protein level in cells from all three germ lineages [22,27]. In iPSCs and the three germ layers, Cx43 protein localized to the cell surface, forming large
Like previous reports, we find that Cx43 is expressed in undifferentiated iPSCs and can be readily detected at the protein level in cells from all three germ lineages [22,27]. In iPSCs and the three germ layers, Cx43 protein localized to the cell surface, forming large gap junction plaques (Figure 2). Despite continued expression throughout differentiation, we find that Cx43 is dispensable for directed germ layer formation. This contrasts with previous studies which report that Cx43 is upregulated during endoderm differentiation and downregulated in ectoderm [22,27]. Indeed, the role of Cx43 in germ lineage specification includes several contrasting reports. Peng et al. 2019 demonstrated that shRNA knockdown of Cx43 has no impact on endoderm formation, as the knockdown cells were able to express definitive endoderm markers SOX17, FOXA2, and CXCR4 [22]. On the other hand, Yang et al. 2019 demonstrate that siRNA-mediated knockdown impedes definitive endoderm formation from human ESCs. Using CRISPR-Cas9 gene ablation, we find that Cx43 knockout iPSCs readily differentiate into all three germ lineages, with no significant difference from wildtype control cells. These discrepancies between our study and the previously published reports might result from our use of CRISPR-Cas9 gene editing rather than RNA interference knockdown methods. However, if that is the case, one would postulate that our knockout cells would exhibit a greater effect compared to previous knockdown studies. Another possibility is that other connexin isoforms could be upregulated to compensate for the loss of Cx43. As reviewed in Figure 9, few studies have investigated the presence of Cx proteins in iPSCs, and thereby the description of established GJIC in iPSCs remains limited. Gap junction coupling is commonly evaluated using small dyes and molecules, such as Lucifer yellow and neurobiotin [28,46]. Passage of these dyes and molecules is largely governed by channel composition, as it determines pore size, channel shape, and voltage gating. As such, various tracer dyes have been reported to be more permissible to Cx43 channels than Cx40 or Cx45 [47,48]. To broaden our investigation, we chose to use tracer dyes, Lucifer yellow and neurobiotin, as these molecules differ in their molecular weight and charge. Our screen of 10 other connexin isoforms, combined with our dye transfer experiments indicate that there is no GJIC compensation happening in our system. Future studies could make use of electrophysiological methods as a more definitive measure of gap junction coupling in Cx43 knockout iPSCs. Furthermore, future studies will determine whether the expression of any of the other Cxs not examined here are altered in our Cx43 CRISPR-ablated iPSCs. Finally, there is a growing body of work surrounding internally transcribed Cx43 isoforms [49]. Our CRISPR gRNAs target the first two thirds of the GJA1 transcript, leaving the possibility of an internally transcribed Cx43 isoform which may continue to perform non-junctional functions within the cell.
While Cx43 has been well studied, there remains much to learn about the other 20 human Cx isoforms, especially given the link of certain Cxs to developmental processes and disease. For example, Cx40 mutations are associated with atrial fibrillation [50] and loss of functional Cx45 channels results in abnormal cardiomyocyte contraction and improper cardiac development [18]. Cx26 mutations are responsible for nearly half of all hereditary deafness cases [51], highlighting the importance of gap junctions for tissue function. To date, limited roles for connexins in stem cell fate decisions have been described. Overexpression of Cx43 in human ESCs positively mediates cardiac differentiation, while mesenchymal stem cells deficient in Cx43 exhibit decreased osteoblast differentiation potential [52,53]. Conversely, expression of Cx32 markedly decreases in the early stages of adipose-derived stem cell differentiation [54]. Interestingly, overexpression of Cx32 and knockdown of Cx43 promotes hepatocyte differentiation, highlighting the dynamic relationship between GJIC and cell fate specification [26]. Therefore, it is apparent that Cxs are important for the maintenance and downstream specification of somatic stem cell populations. Given that many Cxs isoforms are associated with human disease, understanding how they work at the cellular level will provide necessary insights into disease pathology but also reveal how they contribute to cellular differentiation and tissue patterning during development. Figure 9. Human connexin expression profiles described in literature and this study. Human pluripotent stem cells express Cx40.1 [21] and Cx50 [21]. Cx23 expression in pluripotent stem cells and the three germ layers has not yet been characterized. Our study shows control iPSCs and the three germ layers express mRNA transcripts for Cx26 [21,38,51], Cx30.3 [7,21,38], Cx31 [7,21,38], Cx31.1 [7,21,38], Cx37 [7,21,38], Cx40 [7,21,38], Cx43 [7,21,22,27,38,[43][44][45], Cx45 [7,21,38,45], and Cx62 [7,21,38]. Significant increases were viewed for mRNA transcripts of Cx62 in ectoderm; Cx45 in ectoderm and mesoderm; and for Cx30.3, Cx31, Cx32, Cx36, Cx37, and Cx40 in endo-derm. Transcripts for Cx32 and Cx36 were not detected in mesoderm. Implantation of the human embryo is followed by dramatic spatial rearrangements, such as those viewed during gastrulation. Large spatial rearrangement of post-implantation mouse embryos has been shown to be driven by positional cues passed between the embryonic epiblast and the extraembryonic trophectoderm [55,56]. Introducing newer, more sophisticated model systems into future studies would give a more accurate investigation of signaling processes occurring in vivo. The addition of blastoid and embryoid models would complement this study, through the inclusion of influential trophectoderm specific Cx-mediated signals [57,58]. Furthermore, models, such as gastruloids, would add spatioaxial signals possibly lacking in our current embryoid body models [59].
Several reports dispute the role of Cxs in the establishment, survival, and maintenance of human PSCs [7,20,22,29,38,[60][61][62]. For example, dye transfer assays demonstrate the reestablishment of GJIC during iPSC reprogramming [62], and siRNA knockdown of either Cx43 or Cx45 has been shown to negatively impact iPSC reprogramming efficiency [38,60]. Furthermore, ectopic expression of Cx43 and Cx45 enhances iPSC reprogramming efficiency [29,38,60,63,64]. However, while broad-spectrum GJIC pharmacological inhibition via carbenoxolone kills human iPSCs, it appears that this occurs independently from Cx43 as our GJA1-/-iPSCs exhibit comparable morphology and survival in culture relative to control iPSCs (Figure 3) [20]. Comprehensive evaluation of how different Cx isoforms contribute to human PSC survival and potency will help resolve these discrepancies. Additional studies using precise genetic ablation of various connexin isoforms, rather than broad pharmacological blockade, will clarify the role played by Cxs in regulating the self-renewal and differentiation capacity of human pluripotent stem cells.

Conclusions
In conclusion, we present the differential transcript expression for 11 of the 21 Cx isoforms in iPSCs and the three germ layers. Despite significant upregulation of several Cx transcripts in various germ lineages, along with significant upregulation of Cx43 protein in mesoderm, we conclude that Cx43-mediated GJIC is dispensable during lineage specification of human iPSCs.

Data Availability Statement:
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.