Cornelia de Lange Syndrome mutations in SMC1A cause cohesion defects in yeast

Abstract Cornelia de Lange Syndrome (CdLS) is a developmental disorder characterized by limb truncations, craniofacial abnormalities, and cognitive delays. CdLS is caused mainly by mutations in genes encoding subunits or regulators of the cohesin complex. Cohesin plays 2 distinct roles in chromosome dynamics as follows: it promotes looping, organization, and compaction of individual chromosomes, and it holds newly replicated sister chromatids together until cell division. CdLS-associated mutations result in altered gene expression likely by affecting chromosome architecture. Whether CdLS mutations cause phenotypes through impact on sister chromatid cohesion is less clear. Here, we show that CdLS-associated mutations introduced into the SMC1A gene of budding yeast had measurable impacts on sister chromatid cohesion, mitotic progression, and DNA damage sensitivity. These data suggest that sister chromatid cohesion-related defects may contribute to phenotypes seen in CdLS affected individuals.


Introduction
Cohesin, a multi-subunit protein complex, was originally characterized for its ability to tether sister chromatids together from the time they are made during DNA replication until chromosome segregation at anaphase (reviewed in Parenti and Kaiser (Morales and Losada 2018)).Cohesion between sister chromatids, also referred to as trans-cohesion, promotes their accurate segregation during cell division and is necessary for the efficient repair of certain types of DNA damage.Insufficient inter-sister or transcohesion can lead to cell cycle arrest and/or mitotic delays due to activation of the spindle checkpoint, a surveillance mechanism that is critically dependent on cohesion between sister chromatids (Sonoda et al. 2001;Deehan and Heald 2004).
In addition to its role in chromosome segregation, the cohesin complex also plays a critical role in gene regulation through its ability to promote intrachromosomal or cis interactions, particularly in higher eukaryotes.Cohesin promotes the formation of chromatin loops, resulting in chromosome folding and compaction, and thus the formation of topologically associated domains (Gassler et al. 2017;Rao et al. 2017;Schwarzer et al. 2017;Wutz et al. 2017), reviewed in Davidson and Peters (2021).The limits of many cohesindependent loops are defined by chromatin binding of the transcriptional insulator protein CTCF (Parelho et al. 2008;Rubio et al. 2008;Stedman et al. 2008) (reviewed in Merkenschlager and Nora (2016)).Cohesin can also act with the transcriptional coactivator Mediator to bring certain enhancer-promoter pairs into proximity (Kagey et al. 2010), and with CTCF helps to cluster groups of enhancers thereby promoting expression of specific gene sets (Ing-Simmons et al. 2015).Consistent with a role for cohesin in gene regulation, the cohesin loader NIPBL was first discovered for the role of its Drosophila homolog in developmental gene regulation (Rollins et al. 1999) and in mammals heterozygous mutations in the gene result in changes in transcriptional programs (Kawauchi et al. 2009).In fact, some genes depend on cohesin for their expression, perhaps due to the role of cohesin in the spatial organization of genes and enhancers (Seitan et al. 2013;Sofueva et al. 2013;Zuin et al. 2014).Despite this strong correlation between cohesin, chromosome topology, and gene regulation, some cohesin inactivation studies suggest that cohesin may primarily play an indirect role in gene regulation, perhaps by controlling enhancer clustering (Ing-Simmons et al. 2015;Krivega and Dean 2017;Rao et al. 2017;Thiecke et al. 2020).
The genetic disorder Cornelia de Lange Syndrome (CdLS) is characterized by developmental abnormalities such as limb defects, facial dysmorphism, growth retardation, and cognitive impairments.CdLS (CdLS;OMIM 122470, 300590, 610759, 614701, 300882) is classified as a cohesinopathy, a series of related disorders attributable to primarily to mutations in genes that encode subunits or regulators of the cohesin complex.CdLS-causing mutations were first identified in the NIPBL gene which encodes the Nipped-B-like protein, a subunit of a dimeric complex that loads cohesin onto chromosomes (Krantz et al. 2004;Tonkin et al. 2004).Subsequently, CdLS-causing mutations have been identified in other cohesinrelated genes, such as SMC1A, SMC3, and RAD21, which encode core cohesin subunits, the CDCA5 gene encoding cohesion maintenance factor Sororin, and in HDAC8 which encodes a cohesin deacetylase (Musio et al. 2006;Deardorff et al. 2007;Deardorff et al. 2012;Mannini et al. 2013;Kaiser et al. 2014;Boyle et al. 2015).SMC1A and SMC3 are large, coiled-coil-containing proteins that interact to form a heterodimer that interacts with RAD21 and SA subunits, forming the core cohesin complex (Fig. 1a).Most CdLS-causing mutations in genes encoding subunits of the cohesin complex are missense mutations or small in-frame deletions and result in less severe forms of the disorder (Sarogni et al. 2020).
Because cohesin plays multiple critical roles in chromosome structure, function, and segregation, several models have been proposed to explain the phenotypes seen in CdLS.There is compelling evidence that altered gene expression is the central contributing factor to CdLS.CdLS patient-derived cell lines with mutations mapped to NIPBL and SMC1A exhibit altered expression of numerous genes (Liu et al. 2009;Izumi et al. 2015).In addition, in rare CdLS patients whose mutations do not map cohesion genes, and in patients with syndromes related to CdLS, the causative mutations map to a collection of genes involved in gene regulation or global chromatin functions (reviewed in Parenti and Kaiser (2021)).Together, these data strengthen the conclusion that altered gene regulation is a common contributor to this class of developmental disorders.Consistent with this, in vitro studies of the impact of CdLS-associated mutations in SMC1A and NIBPL show reductions in DNA loop extruding activity (Bauer et al. 2021;Panarotto et al. 2022).In yeast, cohesin also plays important roles in rDNA condensation and nucleolar function (Lavoie et al. 2004;Harris et al. 2014).This has led to the proposal that cohesin-dependent changes in the integrity of rDNA repeats, rDNA condensation, and nucleolar structure are causative in cohesinopathies (McNairn and Gerton 2008a;McNairn and Gerton 2008b;Gard et al. 2009;Xu et al. 2013;Ball et al. 2014;Harris et al. 2014).
In some cohesinopathies, defects in sister chromatid cohesion present as morphological changes in mitotic chromosomes.For example, in Robert's Syndrome, a cohesinopathy caused by mutations in the gene encoding the cohesin regulator ESCO2, a separation between some sister centromere pairs can be observed in chromosome spreads (Tomkins et al. 1979;Vega et al. 2005;Gordillo et al. 2008;Vega et al. 2010).However, consistent defects in trans-cohesion have not been reported in CdLS.Some patientderived cell lines with mutations in the cohesin loader NIPBL revealed loss of sister chromatid cohesion phenotypes, but evaluations of patient-derived cell lines with SMC1A mutations and lymphocytes from CdLS patients reported no obvious sister cohesion defects (Castronovo et al. 2009;Revenkova et al. 2009).Similarly, Drosophila models heterozygous for mutations in NIPBL (as is seen in some CdLS patients) show no obvious chromosome cohesion defects (Rollins et al. 2004).Since severe loss in sister chromatid cohesion is lethal (Nasmyth and Haering 2009;Brooker and Berkowitz 2014), it is likely that sister chromatid cohesion defects due to CdLS mutations are too subtle to be detected in assays based on chromosome morphology.
The cohesin complex is highly conserved throughout eukaryotic phylogeny.This allowed us to assess the impact of CdLS mutations on trans-cohesin function using the fungal model, Saccharomyces cerevisiae.Using both sensitized assays and live cell imaging, we characterized the phenotypes of 6 CdLS mutations in conserved residues in the SMC1A protein.Most of the mutations resulted in changes that are consistent with mild defects in trans-cohesion, including reduced chromosome segregation fidelity, reduced sister chromatid cohesion, and spindle checkpoint activation resulting in mitotic delays.These data suggest that some phenotypes in affected individuals may be attributable to subtle defects in sister chromatid cohesion.

Strains and primers
The genotypes of the strains used in this work are shown in Supplementary Tables 1 (diploid strains) and 2 (haploid strains).All strains are derived from YNN281.Yeast media, culture techniques, and strain construction methods are as described in Amberg et al. (2005).

Strain construction
Polymerase chain reaction (PCR)-based methods were used to create deletions of open reading frames and mutated versions of SMC1.Primers used to build and confirm the strains are shown in Supplementary Tables 3 and 4, respectively.Strains containing mutations in the SMC1 gene were built by a modified 2-step gene replacement method.Briefly, SMC1 gene fragments were amplified using high-fidelity phusion (New England Biolabs) polymerase with primer sets (Supplementary Table 3) to generate the appropriate mutations within the gene fragments, which were then cloned into the pRS404 vector (Sikorski and Hieter 1989).The vector was linearized and transformed into the desired yeast strain and colonies were selected on plates lacking tryptophan (the selectable marker in pRS404).After integration of the plasmid was confirmed by PCR, the strains were plated on media containing 5-fluoroanthranilic acid to select against vector sequence (Toyn et al. 2000).Loss of vector was confirmed by PCR, and retention of the mutations was screened for by genomic PCR and confirmed by sequencing.For in vivo sister chromatid cohesion assays, chromosome III was marked near its centromere (coordinates 113101-113583) by integration of the plasmid OPL509 which carries approximately 256 lac operon operator (lacO) repeats (Straight et al. 1996).Correct integration was confirmed genetically.GFP-lacI was expressed from the CYC1 promoter to generate a GFP focus at CEN3.For immunoblot assays, the 6HA-KANMX4 gene fragment was PCR amplified from pYM14 (Janke et al. 2004) with primer sets (Supplementary Table 3) that target the C terminus of SMC1 gene.The PCR product was transformed into the desired yeast strain, and colonies were selected on plates containing G418.The integration of the cassette was confirmed by PCR.

Western blot analysis of SMC1 expression
Total protein extracts were prepared by TCA precipitation (Foiani et al. 1994).Briefly, 5 × 10 7 /ml cells in logarithmic growth were collected and washed once with water and resuspended in 16.6% trichloroacetic acid at room temperature.After addition of the same volume of 0.5-mm zirconia/silica beads (Biospec Products), cells were disrupted by a Bullet Blender for 5 minutes.Beads were washed once with 100 µl of 16.6% trichloroacetic acid, and the resulting extract was spun for 5 minutes at 13,000 rpm at 4°C.The pellet was washed once with acetone and resuspended in 100 µl of SDS-PAGE sample buffer, pH neutralized by addition of 1-M Tris base, boiled for 5 minutes, and clarified by centrifugation.Proteins were separated by SDS-PAGE and detected by immunoblotting, using monoclonal antibodies against HA (C29F4, Cell Signaling) and PGK (22C5, Molecular Probes).

Chromosome segregation assay
Strains were grown in YPAD (yeast extract, peptone, adenine, dextrose) medium at 30°C, except DJC5−.This strain was grown in a medium to select for cells that carried the mini-chromosome used to assay mis-segregation.Cells were diluted and plated on a synthetic complete medium supplemented with 6-μg/ml adenine.This level of adenine allows growth of the ade2-101 yeast cells that have lost the SUP11 gene marker and permits accumulation of the red pigment used to identify segregation errors.Plates were incubated at 30°C until colonies were large (3-4 days) and further stored at 4°C for a few days to enhance differentiation of the color phenotypes.The half-sectoring assay was scored as previously described (Hieter et al. 1985).

Cohesion assay
Asynchronous cells were grown to mid-log phase at 30°C in YPAD media, then pelleted and washed with H 2 O. Then cells were resuspended at 10 −6 /ml in a complete medium (Sunrise Science Products Cat. 1729-500) with 3 × 10 −6 M α factor (GenScript RP01002) and incubated at 23°C for 3 hours.G1 arrested cells were washed 3 times in YPAD with 0.1-mg/ml Pronase E (Sigma P6911), followed by 1 more wash in YPAD.Cells were then resuspended at 5 × 10 6 /ml in YPAD containing 1% DMSO and 15-µg/ml nocodazole (Sigma) and incubated at 23°C for 3 hours to arrest in M phase.A total of 7.5-µg/ml nocodazole was re-added at 1.5 hours.Samples were collected at 0 and 3 hours following addition of nocodazole for scoring.In large-budded cells with 1 DAPI mass and elongated, or separated red dots that were less than 1 µm apart, we scored the number of GFP dots in each DAPI mass.We then plotted the percentage of cells with 2 GFP dots.At least 100 cells were counted in each replicate.Three or more replicates were performed for each strain.

Mitotic delay assay
G1 arrest was achieved as for cohesion defect assay, except that cells were grown in 30°C.G1 arrested cells were washed 3 times in a complete medium (Sunrise Science Products Cat. 1729-500) with 0.1-mg/ml Pronase E (Sigma P6911), followed by washing once in a complete medium.Cells were then resuspended at 5 × 10 6 /ml in a complete medium and incubated in 30°C for 3 hours in CellASIC microfluidics plates (Millipore) for live cell imaging.Spc42-RFP allowed visualization of spindle pole bodies.The time interval between when the single RFP focus split into 2 foci (spindle formation) and when the 2 foci dramatically increased in their separation (anaphase onset) was measured and plotted.

X-ray irradiation sensitivity test
Asynchronous cells were grown to mid-log phase at 30°C in YPAD media, collected by brief centrifugation, washed with water, and resuspended in the original volume of water.The cells were exposed to an RS-2000 source for the designated dosages.Appropriate dilutions of cells were plated on YPAD agar plates and incubated for 2 days.Each experiment was performed 3 times.The average of 3 replicate platings was plotted.

rDNA condensation assay
Log-phase cells expressing the rDNA-binding protein Net1 fused to GFP (Net1-GFP) were suspended at 1 × 10 6 /ml in a SD Sunrise medium (Sunrise Science Products) containing 3 × 10 6 M α factor, for 2 hours at 30°C.The cells were then washed with 3 times with YPAD containing 0.1-mg/ml pronase E (Sigma), followed by 1 wash with YPAD.Cells were then resuspended in YPD containing 1% DMSO at 4 × 10 6 /ml.After 1 hour of incubation at 30°C, rDNA condensation was analyzed in cells with large buds (daughter cell diameter greater than one-half of mother cell diameter) that contained 1 DAPI staining mass by visualizing Net1-GFP.Cells were placed in 1 of 4 categories based on Net-GFP morphology as described previously (Lavoie et al. 2004).The 4 categories were as follows: puff/amorphous (a single large amorphous cap on the edge of the nucleus larger in diameter than one-fourth the diameter of the DAPI staining mass), cluster (a GFP signal overlapping the nucleus, smaller than one quarter the diameter of the DAPI staining mass), loop (condensed arc shape protruding from the DAPI staining mass), or line (a single line of GFP signal extending out from the DAPI mass).

Statistics and sequence analysis
Graphing and statistical analysis were done using Prism software and web-based tools (QuickCalcs), both from Graphpad (La Jolla, CA).Protein alignments and sequence logos were generated using the Clustal Omega algorithm in Geneious Prime (Biomatters, Auckland, New Zealand).Parallel coiled-coil folds were predicted using Paircoil2 (McDonnell et al. 2006).

Microscopy
rDNA assay images were collected using a Zeiss AxioImager microscope with band-pass emission filters, a Roper HQ2 charge coupled device, and AxioVision software.Mitotic delay experiments (every 3 minutes for 3 hours) were performed with Onix2 microfluidics system (Millipore) using Y04C-02 plates with a flow rate of 2 pounds per square inch.Images were collected with a Nikon Ti2 inverted microscope equipped with the Perfect Focus system, an ORCA FLASH camera, automated stage, a Lumencor LED light source, and NIS software.Images were processed and analyzed using Nikon NIS Elements software.Sister chromatin cohesion assay images were also collected with the Nikon Ti2 inverted microscope system described above, and images were processed using Nikon NIS Elements software.

Results
A number of mutations that cause Cornelia de Lange Syndrome [CDLS2 (MIM: 300590)] have been mapped to the gene encoding the core cohesin subunit, SMC1A, which encodes the SMC1A protein (Musio et al. 2006;Borck et al. 2007;Deardorff et al. 2007).SMC1 Cohesinopathy mutations assayed in yeast | 3 proteins have a well-conserved structure from yeast (Smc1) to humans (SMC1A).The protein folds over on itself at a globular region near its midpoint (the hinge), and the N and C termini come together to form a second globular domain, called the head (Fig. 1a).The SMC3 protein is similarly folded, and SMC1A and SMC3 interact at both the hinge region and at their head domains, which form a pair of ATP-binding sites (Fig. 1a).The head domains also interact with the RAD21 subunit (Scc1 in yeast), and SA1 or SA2 (Scc3 in yeast) to form a tetramer.Chromatin-bound cohesin also interacts with a regulatory subunit called PDS5.The long coiled-coils in SMC family proteins are predicted to be interrupted by short regions that extend from the coiled-coils as loops, designated LI, LII, and LIII (Fig. 1b and Supplementary Fig. 1) (Beasley et al. 2002).More recently, these regions have been referred to as the elbow (amino acids 355-410 and 780-815 in human SMC1A) and joint (amino acids 205-255 and 940-1000) regions (Bürmann et al. 2018).The SMC1A and SMC3 proteins can fold over at the elbow regions bringing the hinge in proximity of the joint.In the folded configuration, the SMC1A/3 hinge interacts with NIPBL (called Scc2 in yeast) or PDS5, thereby affecting cohesin loading or DNA loop extrusion, respectively (Bürmann et al. 2018;Petela et al. 2021).Consistent with this, in budding yeast, mutations in Smc1 moiety of the hinge (smc1-D588Y) impact cohesin-Scc2 interactions and mutations in the joint region (smc1-209L1-2) allow cohesin loading but ablate sister chromatid cohesion (Milutinovich et al. 2007;Petela et al. 2021).
The positions of CdLS-associated SMC1A missense and deletion mutations reported in the Human Genome Mutation Database (http://www.hgmd.cf.ac.uk/ac/index.php)are shown in Fig. 1b.We sought to explore the ways in which CdLS mutations might disrupt cohesin function by exploiting simple quantitative assays for many aspects of mainly trans-cohesion function that can be performed in yeast.We chose for analysis 6 mutations that were previously characterized and are associated with CdLS phenotypes of varying severity (Table 1) (Deardorff et al. 2007).The CdLS mutations selected for this analysis were among those that lie in conserved sequence blocks and affect amino acids that are similar or identical in the human and yeast proteins (Fig. 2).Importantly, not all CdLS mutations alter amino acids that are conserved between humans and yeast, and these mutations could impact functions that are not conserved in yeast.For each of the chosen mutations, we created the same mutation in the yeast SMC1 gene at its endogenous locus.

Some conserved CdLS-associated mutation in SMC1 disrupt rDNA packaging
A previous study that used budding yeast to evaluate conserved CdLS mutations in cohesin and cohesin-regulatory genes ECO1 (W216G), SMC1 (E508A; Q843Δ), and SCC2 (R716L; D730V; G1242R) concluded that these mutations did not result in cohesion defects, but 2 of the alleles (eco1-W216G and scc2-D730V) resulted in altered nucleolar morphology and condensation (Gard et al. 2009), while others have shown loss of cohesion in the case of Eco1 W216G (Borrie et al. 2017) We therefore tested for these defects in strains bearing CdLS-associated SMC1 mutations.Cells expressing Net1-GFP were imaged to reveal nucleolar morphology.In wild-type cells, the rDNA in most cells condenses to form a tight loop or line that can be visualized by the associated Net1-GFP (Lavoie et al. 2004).Cells were arrested in G1 with alpha factor then released into the cell cycle.Mitotic cells (large bud, single DAPI-staining nuclear mass) were imaged.Approximately 95% of wild-type control cells exhibited condensed lines or loops of rDNA.In contrast, 4 of the 6 mutants exhibited elevated levels of the less condensed rDNA that manifests as a puff-like Net1-GFP morphology (Fig. 3, a and b).Therefore, like the previously characterized eco1-W216G and scc2-D730V CdLS alleles, these SMC1 CdLS mutations affect rDNA behavior.

Some conserved CdLS-associated mutations in SMC1 result in sensitivity to radiation-induced damage
Cohesin is critical for the timely repair of DNA double-strand breaks, and cell lines derived from CdLS patients carrying mutations in NIPBL (SCC2 in budding yeast), SMC1A, and SMC3 exhibit sensitivity to ionizing radiation (Vrouwe et al. 2007;Revenkova et al. 2009).To test whether CdLS SMC1A mutations affect this cohesin function, we monitored the sensitivity to ionizing radiation

Gender of patient
Yeast mutation of strains bearing the conserved CdLS mutations.Cells in logarithmic growth were subject to 600 Gy of radiation and then plated to assess survival.Two of the 6 mutations (smc1-Δ59-63 and smc1-F1123L) conferred significant sensitivity to ionizing radiation (Fig. 3c).This finding corresponds to results in a study of CdLS patient-derived cells (Revenkova et al. 2009), in which cell lines with cognates of the 2 radiation-sensitive mutations in yeast (SMC1A-ΔV58-R62 and SMC1A-F1122L) similarly exhibit sensitivity to ionizing radiation.In contrast, patient-derived cells with the SMC1A-R496H mutation, like yeast with the homologous smc1-K511H mutation, did not exhibit radiation sensitivity.Thus, in this assay, the yeast and human mutations phenocopy each other.

Some CdLS-associated mutations in yeast SMC1 result in defective chromosome segregation
Prior studies have shown that alterations in gene expression are likely the primary reason that mutations in cohesin-related genes lead to CdLS (Dorsett and Krantz 2009;Liu et al. 2009).In contrast, assays in various CdLS models have not consistently revealed defects in sister chromatid cohesion.However, most of these studies assayed for evidence of profound loss of sister chromatid cohesion -for example frequent and complete separation of sister chromatids in chromosome spreads (Kaur et al. 2005;Castronovo et al. 2009;Revenkova et al. 2009).These studies made clear that gross sister chromatid cohesion failure is a poor diagnostic tool for CdLS (Castronovo et al. 2009).However, those assays were not designed to identify milder defects in sister chromatid cohesion that would be compatible with cell proliferation and organism survival.Indeed, the 6 conserved CdLS mutations evaluated here all support normal growth of budding yeast under non-stressed conditions (Supplementary Fig. 2a), demonstrating that in cells with these mutations, all of the chromosomes segregate correctly in most cell cycles.The mutations did not significantly affect protein expression levels (Supplementary Fig. 2b).Here, we asked whether conserved CdLS-associated mutations in yeast SMC1 might result in mis-segregation of 1 or a few chromosomes in occasional cells, or perhaps put cells under stress during mitosis.To test for this, we used an assay that detects the mis-segregation of a single, truncated, nonessential reporter chromosome (or mini-chromosome).The small size of the test chromosome renders it more sensitive than natural chromosomes to defects in segregation machinery, and thus facilitates the evaluation of mutations that diminish, but do not abolish, segregation fidelity (Hegemann et al. 1988).Mis-segregation of the mini-chromosome provides a colorimetric read-out because the mini-chromosome carries a suppressor of the ade2-101 mutation that causes accumulation of a red pigment (Fig. 4a) (Hieter et al. 1985).The suppressor mutation on the mini-chromosome affects accumulation of the pigment in a dose dependent manner allowing cells that have lost the mini-chromosome, or gained extra copies, to be detected by colony color (Fig. 4a).Because the mini-chromosome carries no essential genes, cells that lose it continue to propagate and can be scored in the colony color assay.In diploid cells, the system allows the detection of both mis-segregation, in which 2 copies of the chromosome segregate to 1 daughter cell and none segregate to the other (2:0 segregation), and chromosome loss, in which 1 copy of the marker chromosome is lost during cell division (1:0 segregation) (Fig. 4a).Chromosome loss may occur when 1 chromatid, which has lost its association with its sister, does not attach to microtubules from either side of the spindle, and is left in the spindle mid-zone at anaphase I, and fails to be included in either daughter nucleus (Hieter et al. 1985).Both types of errors occur when sister chromatid cohesion fails.Thus, this system provides a sensitive assay for defects in establishing or maintaining sister chromatid or trans-cohesion.
Because CdLS-associated SMC1A mutations occur mainly in a heterozygous configuration in females and are present as the sole allele of SMC1A in males, we scored the chromosome missegregation phenotypes of the alleles in both heterozygous and homozygous configurations.When screening the yeast SMC1 Cohesinopathy mutations assayed in yeast | 5 alleles in the heterozygous configuration, we compared their behavior to both the wild-type strain (SMC1/SMC1) and a heterozygous strain, in which 1 copy of SMC1 was deleted from the diploid background (SMC1/smc1Δ).The SMC1/smc1Δ control showed significantly higher chromosome segregation error rates than the wild-type control demonstrating that the SMC1 is haploinsufficient (Fig. 4a).Earlier work in budding yeast has shown that levels of the Scc1/Mcd1 subunit of cohesin can be lowered several fold with no clear impacts on sister chromatid cohesion, although these investigators used different, and probably less sensitive, assays to score cohesin function (Heidinger-Pauli et al. 2010).In the heterozygous configuration, chromosome missegregation was significantly disrupted by only 1 of the CdLS-associated alleles (smc1-Δ59-63) (Fig. 4b and Supplementary Table 2).Because Smc1 is limiting in this assay, an elevated mis-segregation phenotype in a heterozygote could reflect insufficient Smc1 function rather than dominant negative action of the mutant allele.Because none of the CdLS heterozygotes showed a greater defect than the SMC1/smc1Δ control, there is no evidence that any of them have dominant-negative phenotypes in this assay 57 (Fig. 4b).
Five of the CdLS-associated mutations had no significant effect on chromosome segregation when present with a wild-type copy of SMC1 in the cell.Since these heterozygotes (SMC1/smc1-CdLS) exhibited better segregation fidelity than the SMC1/smc1Δ control and indistinguishable from the SMC1/SMC1 control, these mutations must be supplying at least partial function, such that in the presence of a wild-type copy the cells can segregate the test chromosome accurately.To test whether the CdLS-associated mutations lost at least some ability to contribute to chromosome segregation, we re-screened the alleles in the homozygous configuration in diploids (Fig. 4 and Supplementary Table 3).For 2 of the mutants the mini-chromosome was lost so frequently that it was not possible to score the sectoring phenotype (Fig. 4c alleles marked by a red "X").In addition, the smc1-Δ59-63 allele also exhibited severe defects in the homozygous configuration.
A previous sister chromatid cohesion assay of CdLS patientderived cell lines showed that the human cognate mutations of smc1-Δ59-63, smc1-K511H, and smc1-F1123L had no cohesion defects in an assay in which separation of greater than 50% of the sister chromatid pairs in each chromosome spread was scored as loss-of-cohesion (Revenkova et al. 2009).Loss of sister chromatid cohesion at this level is likely to render the yeast strains inviable.Thus, as in the human cells, the CdLS mutations do not catastrophically compromise chromosome segregation in yeast, but the sensitized assay reveals that most of the mutations  (Hieter et al. 1985).In the absence of suppression, ade2-101 cells accumulate a pigmented intermediate in the adenine biosynthetic pathway, resulting in red cells and colonies.The copy number ratio between the marker mini-chromosome and the ade2-101 allele determines the degree of suppression, and thus the colony color.Full suppression (1:1 ratio mini-chromosome:Chr III) results in white colonies, while partial suppression (1:2 ratio, mini-chromosome:Chr III) results in pink colonies, and in the absence of suppression (no mini-chromosome), colonies are red.Colonies are scored based on the first segregation event (half colony only), and smaller sectors that arise later during outgrowth are ignored.Examples of sectored colonies are shown.b) Chromosome mis-segregation rates of strains with heterozygous CdLS mutations (SMC1/smc1 CdLS ).Shown are the percent of colonies exhibiting mis-segregation of the marker chromosome in the first cell division after plating.The total frequency of chromosome mis-segregation events (2:0 segregation and 1:0 segregation) is reported.Symbols: black, wild-type homozygous; gray, wild-type hemizygous; blue, CdLS-associated mutations.c) Chromosome mis-segregation rates of strains with homozygous CdLS mutations (smc1 CdLS /smc1 CdLS ).Red "X" = strains that were not scorable due to extreme instability of the marker chromosome.A statistical analysis was performed with a 1-way ANOVA (*P < 0.05, **P < 0.01, ****P < 0.0001).diminish the ability of Smc1 protein to ensure high fidelity chromosome segregation (Fig. 4a).

CdLS mutations in SMC1 cause cohesion defects
In both human and yeast cells with SMC1 CdLS mutations, chromosomes segregate correctly in most mitoses.However, the results of the sensitized segregation assay (Fig. 4) demonstrate that these mutations compromise the ability of SMC1 to ensure high fidelity chromosome segregation.We therefore performed an assay of yeast strains with CdLS-associated mutations that would allow us to detect mild defects in sister chromatid cohesion.To do this, we tagged a natural chromosome near its centromere with a fluorescent protein, GFP-LacI, that binds to an array of lac operon operator sequence repeats, making a green dot (Straight et al. 1996) (Fig. 5a).These cells were also engineered to express Spc42-DSRed to mark the microtubule organizing center (the spindle pole body: SPB).To perform the assay, haploid cells were synchronized in G1 using alpha factor and released into the cell cycle in the presence of the microtubule depolymerizing agent nocodazole.This treatment causes the cells to arrest in metaphase with a collapsed spindle (yielding side-by-side SPBs).Cells in M phase were identified as those with large buds and a single nucleus.In cells with functional cohesion, the GFP foci on the cohered sisters appear as a single dot in most metaphase cells, whereas reduced cohesion allows the sister chromatids to separate, and 2 dots can be resolved (Fig. 5a) (Michaelis et al. 1997).In the wild-type control (SMC1), 2.97% of the chromosomes exhibited 2 GFP dots (indicating separation of the sister chromatids).In contrast, 4 of the 6 CdLS mutations resulted in significantly elevated levels of separated sister chromatids (Fig. 5b).These 4 mutants (smc1-Δ59-63, smc1-F147V, smc1-K511H, and smc1-F1123L) were also the 4 that showed the highest error rates in the chromosome mis-segregation assay.We conclude that CdLS-associated mutations result in weakly compromised sister chromatid cohesion which is likely the cause of mitotic errors in the sensitized chromosome segregation assay (Fig. 4).

CdLS-associated mutations in budding yeast SMC1 cause mitotic delays
A total loss of sister chromatid cohesion as would be caused, for example, by complete inactivation of SMC1, leads to failed mitoses and is incompatible with cell or organismal viability 49,50 .Four of the CdLS-associated mutations evaluated here cause detectable cohesion defects in 1 chromosome (chromosome III) in a small percentage of cells (Fig. 5).Assuming chromosome III is representative of all 16 chromosomes, then, most cells would be expected to have at least 1 chromosome with compromised cohesion in every mitosis.The high viability of these cells suggests that chromosome segregation is robust.However, reduced cohesion might yet affect cell cycle progression by activation of a checkpoint.The process of correctly attaching sister chromatids to the mitotic spindle requires functional centromeric cohesion (Fig. 6a).Microtubules attach and detach from kinetochores until they reach a stable bi-oriented configuration.Correct attachments are stabilized by tension that is critically dependent on centromeric cohesion between sister chromatids.The spindle checkpoint prolongs metaphase until all sister chromatid pairs are bi-oriented on the spindle.To determine whether the mild cohesion defects caused by the CdLS mutations resulted in spindle checkpoint activation, we assayed our strains for delays in metaphase using live cell imaging.Cells were propagated and imaged in a microfluidics chamber.The SPBs were marked by Spc42-DSRed and images were acquired every 3 minutes to track spindle growth.Cells were scored as entering mitosis in the first imaging frame with 2 SPBs-indicative of spindle formation (Fig. 6b).Entry into anaphase was marked by a rapid increase in spindle length (Fig. 6b).Examples of 10 wild-type and 10 mutant cells (smc1-Δ59-63) from 1 replicate of the experiment are shown in Fig. 6c.Wild-type cells spent an average of 26 minutes in metaphase (Fig. 6d and Supplementary Fig. 3).All the mutants exhibited longer than average metaphase durations, with 2 rising to the level of significance (smc1-Δ59-63 and smc1-F1123L).To determine whether the delays due to activation of the spindle checkpoint, we tested whether ablating the spindle checkpoint shortened the delay by deleting the MAD2 checkpoint gene from our strains.Metaphase timing for the wild-type strain was not affected by the MAD2 deletion consistent with published observations (Li and Murray 1991).In contrast, all 6 mutant strains proceeded more quickly through metaphase in the mad2Δ background, 3 rising to the level of statistical significance (Fig. 6e and Supplementary Fig. 3).

Discussion
The precise ways in which cohesin gene mutations lead to CdLS phenotypes are not known, but as described in the Introduction, there is compelling evidence that altered gene expression is a central contributing factor.
Here, we have performed a series of assays in yeast to determine whether CdLS-associated mutations in SMC1A might also impact functions associated with trans-cohesion.We have not Cohesinopathy mutations assayed in yeast | 7 assayed these collection mutations for their effects on genomewide transcription profiles.However, the human version of one of them (smc1-Δ59-63) was recently analyzed in an in vitro chromatin loop-extrusion assay and exhibited quantifiable defects in this activity as well (Bauer et al. 2021).The compelling evidence that CdLS mutations cause transcriptional defects suggests that the mutations we have analyzed are also likely to be defective in transcription control.
Our findings, summarized in Table 2, demonstrate that CdLS mutations impact cohesin's role in multiple aspects of chromatin organization.Our analysis of 6 CdLS-associated mutations in conserved amino acids of SMC1 reveals that all but one (smc1-K801Q) exhibit quantifiable deficiencies in 1 or more processes associated with cohesin function, many of which rely at least in part on transcohesion (DNA repair, sister chromatid cohesion, metaphase progression, and chromosome segregation), although K801Q had a subtle checkpoint-mediated mitotic delay.Previous assays of the effects of CdLS mutations on trans-cohesion monitored more substantial sister chromatid cohesion defects than the assays used here and found no clear defects (Castronovo et al. 2009;Revenkova et al. 2009).Here, we find that several of the CdLS-associated SMC1A mutations that we evaluated cause low levels of chromosome segregation errors, which is consistent with the compatibility of these mutations with life.Importantly, although the sister chromatid cohesion defects associated with the SMC1A mutations were mild, some of the mutations appear to cause metaphase delays in many or even most cell cycles.Defective sister chromatid cohesion at centromeres or adjacent sites triggers spindle checkpoint-mediated delays in metaphase (Skibbens et al. 1999;Toyoda et al. 2002;Toyoda and Yanagida 2006).This is because stabilization of correct microtubulekinetochore attachments depends on the transmission of tension between the bi-oriented kinetochores, and this requires sister chromatid cohesion at the centromeres.Thus, although the SMC1A mutations do not result in total failed sister chromatid cohesion, we suggest that 1 or a few chromosomes in most cells have deficiencies in generating tension at the kinetochores during metaphase, which then trigger spindle checkpoint-mediated cell cycle delays.These delays in budding yeast cell divisions must not be associated with frequent chromosome segregation errors since the strains exhibit largely normal growth.
In our work, we did not see a clear correlation between the severity of the phenotypes in human patients (Tables 1 and 2) and the assays shown here.The only allele tested that causes a severe phenotype in humans was E493A (E508A in yeast), which only showed a defect in rDNA condensation in the yeast assays.We do note that the CdLS-associated alleles tested here had different impacts in the various assays we performed, consistent with that fact that the different mutations probably affect distinct cohesin interactions, conformations, or activities.Recent in vitro studies showed that charge-altering CdLS mutations in the coiled-coil regions of SMC1A (Fig. 1), for example R790Q (yeast K801Q in our study), diminish SMC1A-SMC3 coiled-coil interactions (the rod conformation) and result in reduced loop extrusion but that these mutations do not seem to impact ATP hydrolysis (Bauer et al. 2021).In contrast, the Δ58-62 mutation (yeast Δ59-63 in our study) removes surface amino acids in the head domain that promote DNA binding.The Δ58-62 mutant shows slightly reduced stimulation of ATP hydrolysis by DNA binding and slightly reduced loop extrusion.Finally, the F1122L mutation maps immediately adjacent to one of the ATP-binding pockets and likely directly impacts the ATPase function (Marcos-Alcalde et al. 2017).Thus, CdLS mutations directly reduce distinct and different molecular actions of cohesin.Combined with our results, it appears that the different functional roles of cohesin (sister chromatid cohesion, chromatin organization and compaction, DNA repair) are differentially impacted when specific molecular activities, such as DNA binding, coiled-coil interactions, or ATPase activity, are reduced.
The 6 mutations we evaluated showed a range of mutant behaviors (Table 2).The K801Q showed no significant defects.Defects in rDNA condensation were exhibited by the other 5, even those with few other defects (F147V and E508A) suggesting that the rDNA condensation assay is perhaps the most sensitive to subtle loss of cohesin activity.It is also possible that the role of cohesin in rDNA condensation in yeast is somehow related to its role in gene expression in human cells (e.g.chromatin loop formation).Chromosome segregation K511H and F1123L, when homozygous, resulted in extreme instability of the marker mini-chromosome.Because the mini-chromosome is so short, it may be especially reliant on centromeric cohesion.It may be that K511, which is not near an ATP-binding site, and F1123 (which is) both contribute to molecular activities that are critical for establishing transcohesion and had significant impact in this assay.Finally, it is also possible that several of the alleles would show statistically significant defects in multiple assays with larger sample sizes.For example, in the chromosome loss and metaphase delay assays, all the alleles trended toward defective behavior but only a subset were significantly different from wild type.It is also possible that some of the assays we chose are not sensitive enough to detect very subtle loss of function.
In metazoans, spindle checkpoint-mediated delays often trigger apoptotic cell death (Ruan et al. 2019).A recent study in Drosophila has shown that mitotic delays triggered by sister chromatid cohesion defects result in developmental defects (Silva et al. 2018).Here, inactivation of a gene (san) that results in modest defects in sister chromatid cohesion caused mitotic delays and severe defects in wing development.These phenotypes could be by-passed by mutations that ablate the spindle checkpoint (mad2 or mps1)-suggesting that the observed developmental defects were attributable to extended metaphase delays.These findings, together with our results, suggest that some CdLS mutations might lead to deficiencies in populating tissues during the appropriate developmental windows due to their triggering of mitotic delays.Thus, although altered gene expression may be the major cause of CdLS, it may be that subtle deficiencies in trans-cohesion or other cohesin functions contribute to developmental outcomes in CdLS patients.

Fig. 1 .
Fig. 1.SMC1A mutations used in this study.a) Cartoon of the cohesin complex.Not all subunits are present at all times.PDS5 can interact in the region occupied by NIPBL in the drawing.Cohesin is a dynamic structure that can open at multiple interaction sites and fold over at the elbow such that the hinge is near RAD21/NIPBL (Bürmann et al. 2018; Petela et al. 2021).b) Locations of the Cornelia de Lange mutations evaluated in this study, numbered 1-6, are shown relative to the estimated positions of coiled-coil domains of human SMC1A protein.Other CdLS-associated missense (red asterisks) or in-frame deletions (blue asterisks) are also shown.Tick marks are at 100 amino acid intervals.Gray shading indicating the probability of coiled-coil formation, calculated using Paircoil2 (McDonnell et al. 2006), is plotted along the length of human SMC1A.Alignments of the predicted coiled-coil domains of human SMC1A and budding yeast Smc1 are shown in Supplementary Fig. 1.

Fig. 2 .
Fig. 2. The sequence context of the human SMC1A CdLS mutations from this study.The positions of residues for the 6 mutations evaluated here (V58-R62del, F133V, E493A, R496A, R790Q, F1123L) are indicated with black arrows.The black bar indicates the in-frame deletion of the first mutation.Clustal Omega alignments of the relevant sequences in the human protein with Smc1 proteins from diverse eukaryotic organisms are shown.The strength of the alignment is indicated by the height of the colored letters at the top.

Fig. 3 .
Fig. 3. Nucleolar condensation and DNA damage repair phenotypes of yeast strains with conserved CdLS-associated mutations.a) Nucleolar condensation was monitored by imaging a GFP-tagged version of Net1.Cells were collected for imaging 1 hour after release from a G1 arrest.Cells with large buds (greater than one-half the diameter of the mother cell-consistent with being in mitosis) were scored for nucleolar morphology (blue, DAPI staining; green, GFP).Most wild-type cells had condensed lines or loops of Net1-GFP signal (top panel).Others had amorphous puffs of Net1-GFP staining (bottom panel).b) Graph showing percent of scored cells with loop or line morphology.Each point represents the average of thirty cells scored in 1 biological replicate.c) Radiation sensitivity phenotypes of conserved CdLS-associated mutations.Cells in mid-log phase were exposed to 600 Gy of X-irradiation then plated for growth on a rich medium.The fraction of surviving cells relative to nonirradiated controls is indicted in the graph.Each point represents 1 biological replicate.For both experiments, a 1-way ANOVA statistical analysis was performed (*P < 0.05, **P < 0.01, ***P < 0.001).

Fig. 4 .
Fig. 4. Chromosome mis-segregation in yeast strains carrying conserved CdLS-associated mutations.a) Illustration of the sensitized mini-chromosome segregation assay.A marker mini-chromosome encodes a suppressor tRNA (SUP11), which suppresses the ocher nonsense mutation in an ade2-101 allele carried on natural chromosome III(Hieter et al. 1985).In the absence of suppression, ade2-101 cells accumulate a pigmented intermediate in the adenine biosynthetic pathway, resulting in red cells and colonies.The copy number ratio between the marker mini-chromosome and the ade2-101 allele determines the degree of suppression, and thus the colony color.Full suppression (1:1 ratio mini-chromosome:Chr III) results in white colonies, while partial suppression (1:2 ratio, mini-chromosome:Chr III) results in pink colonies, and in the absence of suppression (no mini-chromosome), colonies are red.Colonies are scored based on the first segregation event (half colony only), and smaller sectors that arise later during outgrowth are ignored.Examples of sectored colonies are shown.b) Chromosome mis-segregation rates of strains with heterozygous CdLS mutations (SMC1/smc1 CdLS ).Shown are the percent of colonies exhibiting mis-segregation of the marker chromosome in the first cell division after plating.The total frequency of chromosome mis-segregation events (2:0 segregation and 1:0 segregation) is reported.Symbols: black, wild-type homozygous; gray, wild-type hemizygous; blue, CdLS-associated mutations.c) Chromosome mis-segregation rates of strains with homozygous CdLS mutations (smc1 CdLS /smc1 CdLS ).Red "X" = strains that were not scorable due to extreme instability of the marker chromosome.A statistical analysis was performed with a 1-way ANOVA (*P < 0.05, **P < 0.01, ****P < 0.0001).

Fig. 5 .
Fig. 5. Cohesion failure in yeast cells with conserved CdLS-associated mutations.a) Sister chromatid cohesion was scored in a strain with an array of lac operator repeats inserted adjacent to the centromere and expressing a GFP-lacI fusion protein.The spindle pole body is tagged by expression of an Spc42-DsRed fusion protein.Haploid cells were arrested with alpha factor in G1 then released into a medium with nocodazole to prevent spindle formation.Cells with large buds and 2 Spc42-dsRed foci were scored for separation of GFP spots.a) Examples of cells with 2 (top panel) and 1 (bottom panel) GFP spots.b) The percent cells with 2 GFP spots for wild-type controls and strains with CdLS-associated mutations.Each point represents 1 biological replicate of 100-to-140 cells.A statistical analysis was performed with a 1-way ANOVA (*P < 0.05, ****P < 0.0001).

Fig. 6 .
Fig. 6.CdLS-associated mutations in yeast SMC1 cause delays in cell cycle progression.Haploid cells expressing Spc42-DsRed were arrested in G1 using alpha factor then released into the cell cycle.Cells were mounted in a microfluidics chamber and imaged every 3 minutes.The time between spindle formation (the first frame with 2 red foci) and anaphase (the inflection point when spindle length rapidly increases) was measured.a) Cohesin is needed for correct attachment of sister chromatids to the mitotic spindle.b) Cartoon of cells proceeding through metaphase (imaging allowed visualization only of cell shape and the DsRed signal).c) Traces of the distances between the Spc42-DsRed foci in the first 10 individual wild-type cells and smc1-Δ59-63 cells scored in 1 biological replicate.d) Metaphase duration for wild-type controls and strains with CdLS-associated mutations in SMC1.Each point represents the average metaphase duration for the cells measured in 1 biological replicate.A statistical analysis was performed with a 1-way ANOVA (*P < 0.05, ****P < 0.0001).The red dotted line in panels d and e corresponds to metaphase duration of the wild-type control in panel d, to allow comparisons between the 2 panels.e) Metaphase duration for controls and strains with CdLS-associated mutations in SMC1, all deleted for the MAD2 spindle checkpoint gene.Each point represents 1 biological replicate.For each SMC1 mutant the difference in timing, with and without the MAD2 deletion was compared with Student's t-test (*P < 0.05, **P < 0.01).The numbers of biological replicates, the number of cells measured in each replicate, and the metaphase duration for each individual cell measured are shown in Supplementary Fig. 3.

Table 1 .
Mutations in SMC1A assessed in this study.