| Aspect | Evidence-based statement | Key source (author-year) | URL | Publication date |
|---|---|---|---|---|
| Target identity | In the correct T4 context, **agt** is the gene for the **α-glucosyltransferase (AGT)**, distinct from **bgt**, which encodes the **β-glucosyltransferase (BGT)** acting on glucosylated 5hmC/hmC-containing T-even phage DNA; recent work explicitly states agt and bgt encode the α- and β-glucosyltransferases, respectively. (pqac-00000007, pqac-00000009) | Gomez & Waters 2024; Moréra et al. 1999 | https://doi.org/10.1128/jb.00143-24 ; https://doi.org/10.1006/jmbi.1999.3094 | 2024-09; 1999-09 |
| Enzymatic reaction | By UniProt annotation, AGT is **DNA alpha-glucosyltransferase, EC 2.4.1.26**. In the T4 DNA-glucosylation pathway, the analogous BGT reaction is explicitly documented as **UDP-glucose + 5-HMC-DNA → glucosyl-HMC-DNA + UDP**, supporting AGT as the enzyme that transfers glucose from UDP-glucose to 5-hydroxymethylcytosine in DNA but forming the **α** linkage rather than the β linkage. (pqac-00000009, pqac-00000015) | Moréra et al. 1999 | https://doi.org/10.1006/jmbi.1999.3094 | 1999-09 |
| Acceptor substrate | The T4 glucosyltransferases act on **5-hydroxymethylcytosine (5-HMC/5hmdC) already incorporated into duplex DNA**, not on free nucleotide before DNA synthesis; post-incorporation glycosylation yields **5-α/β-glycosylhydroxymethyl-2′-deoxycytidines**. (pqac-00000003, pqac-00000009) | Pozhydaieva 2024; Moréra et al. 1999 | https://doi.org/10.17192/z2024.0117 ; https://doi.org/10.1006/jmbi.1999.3094 | 2024-06; 1999-09 |
| Donor substrate | The experimentally established sugar donor for T4 DNA glucosyltransferases is **UDP-glucose (UDP-Glc)**; structural and biochemical work on BGT shows glucose transfer with **UDP** release, and Pyle 2024 shows T4 βGT bound to **UDP-Glc** in structural analysis. (pqac-00000011, pqac-00000015, pqac-00000000) | Moréra et al. 1999; Pyle et al. 2024 | https://doi.org/10.1006/jmbi.1999.3094 ; https://doi.org/10.1101/2023.12.21.572611 | 1999-09; 2024-12 |
| Stereochemical specificity | T4 encodes **two** DNA glucosyltransferases with different stereochemical outcomes: **AGT forms α-glycosidic linkages** and **BGT forms β-glycosidic linkages** on 5-HMC in DNA. This is the key functional distinction between agt and bgt. (pqac-00000009, pqac-00000007) | Moréra et al. 1999; Gomez & Waters 2024 | https://doi.org/10.1006/jmbi.1999.3094 ; https://doi.org/10.1128/jb.00143-24 | 1999-09; 2024-09 |
| Pathway position | T4 first hydroxymethylates dCMP via **gp42**, then phosphorylates 5hmdCMP via **gp1** to 5hmdCTP, incorporates 5hmdC during replication, and only **after DNA incorporation** do the α/β glucosyltransferases convert it to glucosylated 5hmdC. (pqac-00000003, pqac-00000002) | Pozhydaieva 2024 | https://doi.org/10.17192/z2024.0117 ; https://doi.org/10.1101/2024.01.28.577628 | 2024-06; 2024-01 |
| Biological role | DNA glucosylation is a phage counter-defense that helps T4 DNA evade host nucleases/restriction systems; recent studies show a tradeoff because phages that lose agt and/or bgt escape some defenses targeting glucosylated DNA but become susceptible to systems recognizing unglucosylated 5hmC. (pqac-00000002, pqac-00000007) | Pozhydaieva et al. 2024; Gomez & Waters 2024 | https://doi.org/10.1101/2024.01.28.577628 ; https://doi.org/10.1128/jb.00143-24 | 2024-01; 2024-09 |
| Recent mechanistic perspective | Recent comparative work places T4 βGT among phage DNA-hypermodifying glycosyltransferases and shows **UDP-Glc binding** plus a **flipped-base** configuration in structural representations (PDB 1SXP), reinforcing a base-access mechanism likely relevant to T4 DNA glucosyltransferases generally. (pqac-00000000, pqac-00000010) | Pyle et al. 2024 | https://doi.org/10.1101/2023.12.21.572611 | 2024-12 |
| Structural fold/mechanism (inferred mainly from BGT) | T4 BGT is a **GT-B fold** enzyme with donor nucleotide-sugar binding mainly in the **C-terminal domain** and acceptor DNA/base binding mainly in the **N-terminal domain**; mechanistic work supports a **direct displacement/inverting** mechanism for BGT with **Asp100** as catalytic base. These data are directly for BGT but are useful mechanistic context for annotating the related T4 AGT enzyme family. (pqac-00000014, pqac-00000016, pqac-00000018) | Larivière et al. 2002; Larivière et al. 2003 | https://doi.org/10.1016/S0022-2836(02)01091-4 ; https://doi.org/10.1016/S0022-2836(03)00635-1 | 2002-11; 2003-07 |
| Base-flipping/DNA recognition | Crystal studies of T4 BGT support a **base-flipping mechanism** in which the target site is accessed by flipping the modified base (or abasic surrogate) out of the DNA helix; the enzyme also bends/distorts DNA during recognition. This is the strongest structural precedent for how T4 DNA glucosyltransferases engage duplex DNA. (pqac-00000015, pqac-00000016, pqac-00000018) | Moréra et al. 1999; Larivière & Moréra 2002 | https://doi.org/10.1006/jmbi.1999.3094 ; https://doi.org/10.1016/S0022-2836(02)01091-4 | 1999-09; 2002-11 |
| Quantitative pathway state in wild-type T4 | In wild-type T4 DNA, **>99% of 2′-deoxycytidines** were reported as **5ghmdC** (glucosylated hydroxymethyldeoxycytidine), highlighting how pervasive the glucosylation pathway is in mature T4 genomes. (pqac-00000002) | Pozhydaieva et al. 2024 | https://doi.org/10.1101/2024.01.28.577628 | 2024-01 |
| Quantitative perturbation data | When NgTET was expressed during infection to compete for the 5hmdC substrate, T4 DNA modification shifted from **99% to 55% 5ghmdC**, with increases in **5hmdC to 10.5%**, **5fdC to 2.3%**, **5cadC to 0.9%**, and **unmodified dC to 34.4%**. This experimentally supports that α/β-GT enzymes act on the 5hmdC precursor during infection. (pqac-00000002) | Pozhydaieva et al. 2024 | https://doi.org/10.1101/2024.01.28.577628 | 2024-01 |
| Quantitative enzymology (βGT benchmark) | Recombinant T4 **βGT** shows **apparent Km ≈ 0.41 μM for 5-hmC DNA**, **apparent Km ≈ 16 μM for UDP-glucose**, and **kcat ≈ 77 min−1** on a fully hmC-substituted T4-gt DNA substrate; βGT is **nonprocessive/distributive** and 5-hmC DNA binds about **10-fold stronger** than glucosylated product DNA. These values are for βGT, not AGT, but provide the best available quantitative benchmark for the paired T4 glucosyltransferase system. (pqac-00000012, pqac-00000017) | Terragni et al. 2012 | https://doi.org/10.1021/bi2014739 | 2012-01 |
| Annotation confidence and limitation | Functional assignment of **T4 agt/P04519** as DNA **α-glucosyltransferase** is strong from the T4 genetics/pathway literature, but direct modern structural/kinetic characterization is much richer for **BGT** than for **AGT**; therefore several mechanistic notes above are informed by the homologous T4 βGT and should be treated as informed context rather than AGT-specific proof. (pqac-00000007, pqac-00000009, pqac-00000014) | Gomez & Waters 2024; Moréra et al. 1999; Larivière et al. 2003 | https://doi.org/10.1128/jb.00143-24 ; https://doi.org/10.1006/jmbi.1999.3094 ; https://doi.org/10.1016/S0022-2836(03)00635-1 | 2024-09; 1999-09; 2003-07 |


*Table: This table summarizes evidence-based functional annotation for bacteriophage T4 agt (UniProt P04519), including its reaction, substrates, pathway role, relationship to bgt, and the strongest available quantitative and mechanistic evidence. It is useful for separating direct AGT evidence from broader T4 glucosyltransferase system evidence, especially where βGT is the better-characterized homolog.*