This episode reviews Dardas et al. (2025), which identifies bi-allelic UGGT1 variants in 15 affected individuals as the cause of a distinct congenital disorder of glycosylation (UGGT1-CDG), describes the clinical spectrum, and dissects diverse molecular mechanisms that impair UGGT1 function.
0:00Welcome to Base by Base. Hello. Today we're embarking on a detailed scientific exploration. We're looking into a newly identified congenital disorder of lycosolation. That's right. And we'll be grounding our discussion in a really significant research paper.
0:14Exactly. It's titled Bioleic UGGT one variants, cause a congenital disorder of glycosolation, and it was published by DARDIS and colleagues in the American Journal of Human Genetics just this year, 2025.
0:26It's a foundational piece. Our aim today is really to understand the, uh, the genetics, the clinical picture and the molecular side of things for this disorder, which they've termed UGGT one CDG. We want to help you grasp the core findings, you know, why this discovery is important.
0:43We'll stick closely to the structure the author is used. So introduction first, then methods, results. And finally, the discussion and conclusions. Sound good. Sounds perfect. Let's dive into the introduction.
0:54The paper starts by setting the stage with glycosylation. Remind us what that is. Sure. So glycosylation. It's a really fundamental biological process. It's basically about attaching these sugar chains, called glycans, onto proteins or sometimes lipids.
1:09Seems simple, but it's crucial, right? Absolutely essential. These sugar editions change how proteins function, how they interact, where they go in the cell. It affects a huge range of things. And when this process has errors, that's where we get into congenital disorders of glycosillation or CDGs.
1:24Exactly. And CDGs aren't just one thing. It's a large, very diverse group of metabolic conditions, all stemming from defects somewhere in that glycosylation pathway. It's a big umbrella term then. Very much so.
1:36The paper then zooms in on N-linked glycosylation. specifically the quality control aspect in the endoplasmic reticulum, the ER. It mentions chaperoned proteins, Calmexon, and Cal reticulin. Why is that quality control steps so important?
1:50Well, think of the ER as a protein folding factory. Cal Nixon, or CNX, and Cowarticulin, CRT, are like the quality control inspectors. They help newly made proteins fold into their correct three-dimensional shape.
2:03And if they don't fold correctly. They can't do their job properly, and they can even cause problems like clumping up inside the cell. So this quality control is vital. Okay, and this is where the main focus of the paper comes in UGGT1.
2:14What's its specific job? UGG21. Its full name is UDP dash glucose. glycoprotein, glucocell transfers, one acts as a key sensor in this system. It checks proteins, and if it finds one that's misfolded, it adds a glucose molecule back onto its glycan chain.
2:31The glucose tag. And what does that tag do? That tag is basically a signal. It tells the misfolded protein, go back and try folding again with the help of CNX and CRT. It allows the protein to re-enter that folding cycle.
2:43Like giving you the 2nd chance to fold correctly? Precisely, a molecular do over. The paper also mentions a related protein, UGGT2. How is that different? Yeah, UGGT 2 is a homologue, so it's similar, but it's generally less abundant than UGGT1.
2:58And uh, it seems to prefer different types of proteins as substrates. The thinking is UGGT1 might handle more of the larger proteins destined for secretion while UGGT2 perhaps focuses on others. So UGGT one is really the main player for this pathway they're studying.
3:13It seems to be, yes. And its importance is really underscored by animal studies. Right. The paper mentions that knocking out the UGGT one gene in mice is lethal. That's correct. Mice without functional UGGT1 don't survive embryonic development.
3:27It just highlights how absolutely essential this protein's function is. Which brings us directly to the core of this study by DARDIS and colleagues. They set out to characterize a specific CDG, caused by problems in the human UGGT1G.
3:41Yes, they identified 15 individuals from 10 unrelated families who all had pathogenic variants in both copies of their UGGT1 gene, what we call bioleic variants. And they described a set of characteristic clinical features in these individuals.
3:56What were the main ones? The most consistent findings were developmental delay, intellectual disability, seizures, quite distinct facial features, and microcephaly, which is a smaller than usual head size.
4:07So primarily neurological features. Primarily, yes, but some cases were more severe and involved other systems, too. They observe things like congenital heart malformations, skeletal issues like scoliosis, and problems with the liver, or kidney specifically, polycystic kidneys in some individuals.
4:23And if the molecular level, what did they find was going wrong, because of these UGGTT1 variants? Their lab work showed these genetic changes could mess things up in a few ways. Some variants clearly impaired UGDC one's ability to add that glucose tag its catalytic activity.
4:42Others disrupted how the genes message, the MRNA, was spliced together, and one variant seemed to prevent the protein from staying in the ER where it needs to be. So different ways to break the system, essentially.
4:53Exactly, leading to a loss of UGGT one function one way or another. Okay, that gives us a really good overview from the introduction. Let's move into the methods. How did the researchers actually carry out this study?
5:04They used a combination of approaches? Detailed clinical investigation of the patients and their families alongside, um, quite sophisticated lab experiments to probe the function of the gene and protein.
5:14Makes sense. How did they find the affected individuals to begin with? It a rare disorder. It is. They leverage several networks. Gene Matcher was one that's a platform connecting researchers working on similar rare genes.
5:27Ah, I've heard of that. Very useful. Extremely. Also, the UK's 100,000 Genomes project. collaborations with other research centers, and direct referrals from clinicians who suspected a CDG based on symptoms.
5:40And the families involved came from various backgrounds. Yes, quite diverse origins, Amish, Pakistani, European, Saudi Arabian, Turkish, Egyptian. This diversity is important for, you know, understanding the broader picture of the disorder.
5:54They even use population specific databases, like one for the Anabaptist community, and another hospital-based one in Saudi Arabia, to help interpret the genetic findings. Good point. So once they identified the families, How do they collect the clinical data?
6:09They use standardized forms called performas? filled out by the clinicians involved. This ensured they collected consistent information on symptoms, medical history, development, etc. They also retrospectively reviewed brain MRI scans for 7 of the individuals.
6:24Looking for patterns in the brain imaging. And then the genetics. How did they pinpoint the UGGT one mutations? For the 1st few families, they used whole genome sequencing, which reads pretty much the entire DNA sequence.
6:38For the rest, they used whole XOM sequencing, which focuses just on the protein coding parts of the gene, the XOM. This is often where disease causing mutations are found. Both powerful techniques. But they generate a lot of data.
6:51How do they filter it down to find the specific UGGT one variants? They had a strict filtering strategy. First, looking at data quality, then focusing on variants predicted to actually change the UGG21 protein, like nonsense or frame shift mutations or misense changes in important areas.
7:08They checked if the variants segregated correctly within the family, meaning affected individuals had 2 variants, parents were carriers, et cetera. And critically, they looked at allele frequency. How common the variant is in the general population.
7:20Right. For a rare disorder like this, the causal variants should be very rare or absent in large population databases like Nome D. Anything common was filtered out. And the variants that pass these filters.
7:31Those were then carefully assessed by researchers at different centers to see if they fit the clinical picture. Finally, they used traditional Sanger sequencing to confirm the presence of the identified UGTT one variance in the patients and available family members.
7:44So, a multi-step process to kneel down the genetic cause. A very thorough one, yes. Okay, that covers the clinical and genetic studies. What about the lab experiments designed to test the function of these variants?
7:56The 1st one mentioned is a UGTT1 cellular reglucasylation assay. What was the goal there? The goal was to directly test if the mutant UGGT1 proteins could still perform their main job, adding that glucose tag back onto misfolded proteins inside a cell.
8:11And how do they set that up? They used HEK 293 cells, a common lab cell line. Crucially, they used versions where another gene, ALG 6 was knocked out. Why knockout ALG 6? Knocking out LG 6 changes the structure of the sugar chains on proteins in a way that makes them more dependent on UGGT1 for proper interaction with the Calnex and Cal reticulin cycle.
8:32It basically makes the effect of faulty Uchi GT1 easier to see and measure. Ah it sensitizes the system. Exactly. Then, they introduce plasmins, little circles of DNA carrying instructions to make either the normal wild type, UGGT one, or one of the mutant versions.
8:48They also introduced unknown substrate for UGGT1, a misfolded protein called AATZ. And how did they measure the glucosillation? They use a pull down technique. They incubated cell extracts with cow reticulin attached to beads.
9:03Since cow reticulin binds to glucose tag proteins, They could pull down anything that UGGT1 had successfully tagged. I see. Then they used Western blotting, which detects specific proteins to see how much UGGT1 itself gut tag, again self glucosylate, and how much of the ATZ substrate got tagged by the different UGGT1 versions.
9:22Clever. What about the 2nd assay, the catalytic activity assay? How is that different? This one was done in vitro? Meaning in a test tube, not in cells. They purified the different mutant UGGT1 proteins and directly measured their ability to transfer glucose from a donor molecule, UDP glucose, onto specific fluorescently labeled sugar substrates.
9:41So, a more direct measure of the enzymes raw chemical activity. Precisely. They used HPLC, high-performance liquid chromatography to quantify how much fluorescent substrate got glucosolated by each UGGT1 variant.
9:55Got it. The methods also mention using patient derived cells, fibroblasts. Why do that? Working with skin fiber blasts cultured from the affected individuals provides a more, you know, physiologically relevant context.
10:09You can study the protein as it exists naturally in the patient's own cells looking at its levels, stability, maybe localization. Which leads to the localization studies. What they looking for there? UGT21 is supposed to function inside the ER.
10:22Some mutations, particularly near the end of the protein might disrupt the signal that keeps it there. So they wanted to see if any mutant versions were escaping the ER and being secreted out of the cell instead.
10:32And how do they check that? They express tagged versions of UGGT1 and HEK cells, and then measured UGGT1 levels both inside the cells in the wholesale ice state, and in the culture medium outside the cells.
10:44They also did similar checks using the patient fiber blasts. Makes sense. And the final method was the mini gene splicing asset. What was that designed to test? This was specifically to investigate if some of the variants, particularly misense ones that didn't seem to directly kill enzyme activity, might actually be causing problems at the level of MRNA splicing.
11:03Splicing. That's the process of cutting out non-coding bits from the RNA message before it's translated into protein. Exactly. Sometimes a DNA change can create a wrong signal for splicing, leading to bits being left in or cut out incorrectly, ultimately messing up the final protein.
11:20So how did the mini gene assay work? They created artificial mini genes containing the relevant UGG21 Exxons and surrounding intron sequences, both the normal version and versions with the suspected splicing mutations.
11:33They put these into cells, let the cells process the RNA, and then use deep sequencing to analyze the exact structure of the resulting MRNA transcripts. So they could directly see if the splicing patterns were abnormal for the mutant versions?
11:46Yes, they could pinpoint exactly how the splicing was altered. Okay, that's a really comprehensive toolkit they applied. Let's transition to the results. What did all this work reveal, starting with the genetic findings?
11:57Well, as mentioned, in that 1st Amish infant, they found compound heterozygis variants, 2 different mutations. One was a small deletion causing a premature stop code on P tier 127. Likely leading to a truncated nonfunctional protein.
12:10Almost certainly. The other was an in-frame deletion of several amino acids, P, ATT 390, JLY 39 and 7. And across the other families. Across the 10 families studied with exome or genome sequencing, they found 9 distinct pathogenic UGG21 variants in total.
12:29These included one nonsense, 4 small insertions dilutions in Antalas, and 4 misense variants. A mix of types. And crucially, these were all rare. Extremely rare, or completely absent from the large Noah Madi database which strongly supported their role in this rear disease.
12:45They also checked conservation for the missense variants. Meaning, are those amino acids the same across different species? Right. And they found these occurred at positions that are highly conserved, suggesting they're functionally important.
12:56Plus, prediction tools like Splice AI flagged a couple of the misins variants, P. Alice 7, the one valve, and P saw 1272 his as potentially disrupting splicing. Okay. And did these variants track with the disease in the families?
13:09Yes, co-segregation analysis confirm that affected individuals had 2 variants, while unaffected parents or siblings usually had only one or none. They did note one family, S 11, where a homozygous splice variant in UGGT one was found, but its definite pathogenicity was uncertain because the individual also had a variant in another gene, FCSK.
13:31Ah, complicating factor there. A bit, yes. But a really interesting finding was the recurrent PR 446 nonsense variant. That was the one found in multiple families. Yes, in 4 unrelated families, all with origins in Saudi Arabia or Egypt.
13:47Hapletype analysis looking at surrounding genetic markers, strongly suggested this is an Arab founder variant. Meaning it arose once in an ancestor and spread within that population group. Very important for diagnostics in that region.
13:59Absolutely. All right, let's pivot to the clinical picture that emerged from the 15 affected individuals. You mentioned variability. Yes, the severity was quite variable, but there were those core features we discussed.
14:10Developmental delay, intellectual disability, seizures, the characteristic facial features, and microcephaly. These were pretty consistent. Can you describe the facial features in a bit more detail? They often included things like a small jaw, micronatia, a long face, sometimes described as slightly coarse features, highly arched eyebrows, and a rounded nasal tip, though, like the overall severity, there was a variation here too.
14:34And other combin issues. Seizures were very frequent, occurring in most individuals, and the types varied focal generalized epileptic spasms. Skeletal anomalies were seen in a subset, things like scoliosis or abnormalities of the fingers and toes, and behaviorally, traits associated with autism spectrum disorder were noted in several patients.
14:54What about organ systems beyond the neurological and skeletal? Congenital heart disease was found in 5 individuals about a third. Less commonly, they saw genitorinary anomalies, like cystic kidney disease, and some liver or biliary tract issues.
15:07And the brain imaging. Was there a consistent pattern on MRI? Not entirely consistent. Some showed abnormalities like gray matter heteratopia, meaning clumps of neurons in the wrong place, or issues with myelination, thinning of the corpus callosum, or cerebellar hypoplasia.
15:24But importantly, some individuals had normal brain MRIs. So a normal MRI doesn't rule it out. Correct. EEG, which measures brain electrical activity, was usually abnormal reflecting the seizure propensity.
15:36And a really critical point for diagnosis. What about the standard CDG blood test? The clinical carbohydrate deficient transfer and testing, which is a common screening tool for N-linked CDGs? Was performed on 3 affected individuals, and it came back normal in all of them?
15:52That's huge. I mean, this specific CDG subtype won't be picked up by that standard test. Exactly. It has significant diagnostic implications. Okay, now let's connect this back to the lab work. What did the functional assays show about how these UGGT one variants actually affect the proteins job, the cell-based glucose solations assay first?
16:09That assay showed pretty clearly that variants predicted to cause major protein disruption, like the P tire one to 7 nonsense mutation, the PTFP 390 GLE 397 Dell deletion, and another frame shift, PGLN 1361 Profs 27 led to a significant loss of UGHG1's ability to glucosolate itself, and the AATZ substrate.
16:31Makes sense for those types of mutations. What about mis sense variants? One misence, PAO 7 to Alicer, showed reduced transglucosillation of ATZ, but didn't seem to affect self-glucosillation much. However, several others, including PLS 711 Vault, PR 127 isui his and the recurrent PR 2346 nonsense variant, didn't show a significant impact in this particular cell-based assay.
16:52Hmm, okay. But what about the in vitro catalytic activity asset? Did that provide more clarity? It did for some. That direct enzyme essay showed partial to complete loss of activity for several variants, including PTR 127, PHE 390 DLA 1397 Dell, PT 723 SIS, another Miss S, PTL, and 1155 Arg, and the PTL and 1361 profs, 27 frame shift.
17:13And interestingly, PR 1272, his, which looked okay in the cell essay, showed a significant loss of activity in this direct test. So a direct enzyme essay picked up defects the cell-based one missed in some cases.
17:22Yes. However, even in this assay, PLS 7-Eleven ball and PR 1546 didn't show a significant reduction in activity compared to wild type. So for those two, the problem wasn't necessarily a direct hit to the catalytic function itself.
17:37Right, suggesting other mechanisms were at play. Which brings us to the results for PR 1546, the founder variant. What did they find regarding its localization? Well, first, when they just looked at protein levels inside the cells using Western Blot, most mutant UGGT1 proteins were detected, except for the ones expected to be rapidly degraded, like TR 127, but for PR 1546, they found something striking.
18:01What was that? Cells expressing the PR 1546 variants secreted a significant amount of the UGGT1 protein into the culture medium, much more than cells with normal UGGT1 or other variants. So it was escaping the ER.
18:14Exactly. And consistent with this, when they looked at fiber blasts from patients homazygus for PR 1546, the amount of UGGT1 protein inside the cells was markedly reduced, while substantial amounts were found outside in the medium.
18:26So the loss of function for this variant is due to it not being in the right place, not because the enzyme part is broken. Precisely. It lacks the C terminal or EL sequence, which normally acts as an ER retrieval signal.
18:36Fascinating. And finally, the mini gene splicing results for those other 2 variants, P-Palo 711 vowel and PR 1272 hins. The mini gene asset confirmed that both of these missense variants do indeed cause splicing errors.
18:50PR conolivol actually introduced a new abnormal splice site, resulting in a 7 base paradeletion and a frame shift in the MRNA. Which would likely lead to a nonfunctional protein. Yes, probably degraded by nonsense mediated decay.
19:06And PR 1272 caused complex alternative splicing, leading to transcripts that either skip an exon entirely or retain part of an intron. Also likely leading to nonfunctional products. Correct. So for these 2 variants, the primary defect is at the RNA processing level, not directly at the protein activity level, even though PR 1272 also showed reduced activity in vitro.
19:27Wow, okay. So, multiple ways these mutations can disrupt UGGT1 function catalytic activity, splicing, localization. Let's move into the discussion. How did the authors synthesize all these findings? They firmly establish UGGT1 CDG as a distinct congenital disorder of glycosylation, characterized by this variable multisystem phenotype, but with neurological problems being front and center.
19:51They note the clinical overlap with some other CDGs affecting ER quality control, like Magya, CDG, and MN1 and B1 CDG. And they reiterated the importance of UGGT1's different functional parts. Yes, they emphasize the roles of the different domains, the TRXL domains for recognizing the misfolded substrates, the core glucosill transphrase domain for the actual catalysis, and that C terminal real motif for keeping it in the ER.
20:18And they highlighted that most of the variants they identified ultimately lead to significantly reduced or absent glucosyl transferease activity, whether directly or indirectly. They also address the discrepancies between the different functional assays for some variants, didn't they?
20:32They did. They suggested potential reasons? Like maybe the overexpression in the cellular asset could mask partial defects, or perhaps the way substrates are presented is different in the test tube versus the cell.
20:43It highlights the value of using multiple approaches. And for the splicing variants. They concluded that the splicing defects identified for PLS 711 Vaul and PR 1272 is almost certainly lead to loss of function, likely through nonsense mediated decay of the faulty MRNA.
21:01And for the PR 1546 founder variant. The key takeaway there is that causes reduced intracellular UGGTT1 levels because the protein is abnormally secreted, impairing ER quality control that way. So the overall consequence in UGGT1 CDG is a disruption of that crucial CNXCRT folding cycle.
21:20That seems to be the core problem. Without functional UGGT one putting that glucose tag back on, misfolded glycoproteins can't efficiently re-engage with the folding machinery. This likely leads to an accumulation of misfolded proteins, potentially causing ER stress.
21:34And they speculated that secreted misfolded proteins might also contribute. Yes, particularly for the PRG 1546 variant where UGDT1 itself is secreted. Maybe other misfolded proteins escape too, potentially causing problems outside the cell, but that's more speculative.
21:50Did they see any link between the type of mutation and how severe the disease wise? They proposed a possible correlation. It seems individuals with variants expected to cause a complete loss of function, like nonsense or frame shift mutations, might tend to have more severe presentations compared to those with misence variants that might retain some residual activity hypomorphic, or primarily affect localization.
22:12But more data is needed to confirm strong genotype phenotype correlations. And that finding about normal transferring glycosolation. They really emphasize that point. Normal transferring glycosylation patterns as seen in the tested UGGT1 CDG patients show the limitations of this standard diagnostic test.
22:29It means you can't rely on it to rule out all CDG. Which underscores the need for genetic testing. Absolutely. If the clinical picture suggests a CDG, genetic testing, like XOM or genome sequencing, is crucial, even if the initial blood tests are normal.
22:44This finding has real implications for CDG classification and diagnostic strategies. And regarding treatment. Currently, like most CDGs, treatment is largely supportive, managing seekers, developmental support, et cetera, but getting an early, accurate genetic diagnosis is vital, partly for prognosis and genetic counseling, but also because it opens the door for considering potential future therapies as they emerge.
23:09Did they comment on the long-term outlook? They noted that survival into adulthood is possible for individuals with UGGT1 CDG, but more long-term follow up studies are needed to fully understand the prognosis.
23:20Okay, let's wrap up with the main conclusions from DARDIS and colleagues' paper. The key conclusion is that biologic pathogenic variants in UGGT1 are definitively a cause of a severe multisystem congenital disorder of glycosylation, with neurological dysfunction being the primary impact.
23:36They also highlighted areas for future work. Yes, definitely. More research is needed to fully map out the entire phenotypic spectrum to better understand those genotype phenotype correlations, and importantly, to investigate whether potential treatments, like maybe sugar supplementation, could offer any benefit, though that's still very much an open question.
23:56And the significance of the founder variant. The identification of the PRR 1546 founder variant in Arab populations is a significant finding with direct implications for genetic counseling and diagnostic approaches in those communities.
24:09So overall, this study expands the known landscape of N-linked glycosylation disorders. Significantly so. It provides critical new insights into the essential role of UGG21 in human health and disease.
24:21This research by DARDAS and colleagues really lays the groundwork for future understanding of UGG1 CDG. It certainly does, a really important contribution to the field. This work was based on an open access article.
24:34under a CCBY4 license. That's right. And for anyone wanting to read the original paper, the DOI and license link are available in the description.
