# Avian Reovirus Sigma C Variation Extends Beyond the Receptor-Binding Head

I expected the globular head of Sigma C to dominate the variation map. It did not.

The head contained substantial amino acid diversity, including variation across a published candidate antigenic peptide. But when differences from the avian reovirus S1133 reference were placed onto a trimeric Sigma C model, variation also extended along the stem that supports and presents the head.

That pattern does not prove altered receptor binding, immune escape, trimer stability, or vaccine failure. It does change the experimental question. Sigma C should not be treated only as a receptor-binding head. Its stem is part of the attachment fibre and deserves direct functional study.

![Figure 1. Sigma C sequence curation and reference-anchored clustering workflow.](/images/molecular-biology/avian-reovirus-sigma-c-variation/01_sequence_workflow.png)

*Figure 1. Public Sigma C records were curated to a unique protein set, then a near-full-length subset was assigned to reference-anchored genetic clusters. One ambiguous GC4-like singleton was excluded from the final 904-sequence presentation set.*

## Key findings

- The cluster-level analysis included **904 unique near-full-length Sigma C sequences** assigned to seven reference-anchored genetic clusters.
- Amino acid differences from S1133 occurred across both the C-terminal globular head and the stem.
- The broad all-NCBI comparison showed greater average diversity in the unresolved N-terminal portion and coiled-coil region than in the head.
- A published predicted and ELISA-recognized peptide, S1133 residues 254 to 260 (`THIPSDL`), varied among cluster consensus sequences. D259 was nearly invariant in the local alignment.
- A difference from S1133 could represent a stable cluster signature or polymorphism within a cluster. These patterns have different interpretations.
- The analysis generated testable hypotheses. It did not establish receptor contacts, antibody escape, altered pathogenicity, or vaccine mismatch.

## Why the Sigma C stem matters

Sigma C is the trimeric attachment fibre of avian reovirus. Structural studies showed that its C-terminal receptor-binding fragment contains a globular beta-barrel head and part of a triple beta-spiral shaft.[1] A later structure of residues 117 to 326 extended this architecture to include an alpha-helical triple coiled coil, a zinc-containing linker, the beta spiral, and the head.[2]

The head remains the best-supported receptor-binding region. An isolated C-terminal fragment retained receptor-binding activity, which is direct experimental evidence for that role.[1] However, the available avian reovirus structures do not show Sigma C bound to a confirmed chicken receptor. A conserved pocket or exposed residue in the head may reflect functional or structural constraint, but sequence conservation alone cannot identify a receptor-contact site.

The stem has a different but equally physical role. It holds the three protomers together and positions the head away from the virion surface. Changes within a coiled coil, linker, or beta spiral could plausibly affect packing, flexibility, or head presentation. Those possibilities remain hypotheses until tested with properly folded trimeric proteins or matched viruses.

## Methods

### Sequence collection and curation

NCBI Protein was queried on July 18, 2026, for avian orthoreovirus records annotated as Sigma C, sigmaC, or cell-attachment protein. All 6,330 Avian orthoreovirus Nuccore records in the local snapshot were also downloaded, and their annotated Sigma C coding sequences were inspected. Annotated translations were retained when available; otherwise, matching coding sequences were translated to the first stop codon.

Protein-derived and nucleotide-derived translations were merged. Terminal stops and alignment gaps were removed, exact amino acid duplicates were collapsed, and source accessions were retained in a provenance table. This process produced 1,868 unique Sigma C proteins, including the 326-amino-acid S1133 reference AAK18188.1. The broad positional analysis used the 1,867 non-reference sequences, including partial records, with coverage calculated separately at every S1133 position.

A near-full-length subset contained 1,044 unique proteins between 310 and 330 amino acids. These proteins were used for reference-anchored cluster assignment.

### Reference-anchored cluster assignment

GC1 through GC6 anchors were obtained from the supplementary tables of Egaña-Labrin and colleagues.[4] Eight GC7 anchors were recovered from the Bilal and colleagues phylogeny and verified in GenBank.[5]

Each near-full-length sequence was compared with the recovered anchors by pairwise amino acid p-distance across jointly covered valid residues. A sequence was assigned to the nearest reference cluster when its closest-anchor distance was no greater than 0.45. That cutoff served as an analysis rule rather than a formal taxonomic boundary.

Nine hundred five sequences met the assignment rule. PQ106587.1, an ambiguous long-branch GC4-like singleton, was excluded from the presentation tree and GC4 frequency calculation. The final presentation set therefore contained 904 sequences.

### Tree construction and structural mapping

The 904 retained avian reovirus proteins and a Nelson Bay orthoreovirus outgroup were aligned with MAFFT. Pairwise amino acid p-distances were calculated after gaps and nonstandard residues were excluded for each comparison. A BIONJ distance tree was constructed and rooted with the Nelson Bay sequence. No bootstrap support values were calculated.

At each S1133 position, the proportion of valid residues that differed from S1133 was calculated separately for every cluster. S1133 itself was excluded from the GC1 frequency denominator. Position-specific frequencies were then mapped to the corresponding residues in each protomer of a full-length AlphaFold trimer. Gray represented no observed difference from S1133, and increasing pink-to-red intensity represented a higher within-cluster difference frequency.

The D-to-E connection in the globular head was operationally defined as residues 250 to 273 from the experimentally resolved 2JJL secondary structure.[2] In that structure, residues 243 to 249 and 274 to 282 formed the flanking beta strands, while the intervening segment included two short helices. The candidate peptide THIPSDL, residues 254 to 260, was treated as a separate nested interval. The full-length trimer was rendered in PyMOL as a molecular surface with a mesh overlay.

## Results

### Seven reference-anchored clusters contributed to the map

| Genetic cluster | Unique near-full-length sequences |
| --- | ---: |
| GC1 | 215 |
| GC2 | 270 |
| GC3 | 45 |
| GC4 | 96 |
| GC5 | 185 |
| GC6 | 64 |
| GC7 | 29 |
| **Total** | **904** |

These counts describe unique public protein sequences in the curated snapshot. They are not estimates of prevalence, outbreak frequency, farm-level incidence, or geographic distribution. Exact-sequence deduplication also means that repeated deposits of an identical protein did not increase its representation.

![Figure 2. Reference-anchored cluster composition.](/images/molecular-biology/avian-reovirus-sigma-c-variation/02_cluster_composition.png)

*Figure 2. Unique near-full-length Sigma C sequences assigned to GC1 through GC7. The bars describe the composition of this curated public-sequence snapshot, not field prevalence.*

![Figure 3. Reference-cluster topology.](/images/molecular-biology/avian-reovirus-sigma-c-variation/03_reference_cluster_tree.png)

*Figure 3. Simplified rooted BIONJ topology for the GC1 through GC7 reference anchors, with Nelson Bay orthoreovirus as the outgroup. Branch lengths were omitted for clarity.*

### Diversity was distributed across the fibre

The all-NCBI comparison showed that every S1133 position had at least one observed alternate amino acid among covered sequences. The strongest 20-residue window occurred at positions 94 to 113, within the N-terminal portion that was not included in the experimental 117 to 326 structure.

| S1133 region | Mean non-S1133 frequency | Mean amino acid entropy | Evidence-bounded interpretation |
| --- | ---: | ---: | --- |
| 1 to 116, unresolved N-terminal portion | 0.560 | 1.763 bits | Most diverse broad region; atomic interpretation remained limited. |
| 117 to 156, coiled coil | 0.497 | 1.608 bits | Substantial divergence within an experimentally resolved trimeric element. |
| 157 to 195, beta spiral and linker | 0.362 | 1.178 bits | Less divergent on average, but not invariant. |
| 196 to 326, globular head | 0.370 | 1.124 bits | Variable receptor-binding region with exposed candidate sites. |
| 250 to 273, DE-loop region | 0.414 | 1.147 bits | Structurally defined head interval that included the candidate peptide. |

The averages depend on S1133 as the reference and on the composition of public sequence databases. They do not measure an evolutionary rate or selection.

![Figure 4. Position-specific differences from S1133.](/images/molecular-biology/avian-reovirus-sigma-c-variation/04_position_specific_variability.png)

*Figure 4. At each S1133 position, the value is the proportion of covered public Sigma C sequences carrying a valid amino acid different from S1133. The candidate peptide THIPSDL spans residues 254 to 260.*

![Figure 5. Mean sequence diversity by Sigma C region.](/images/molecular-biology/avian-reovirus-sigma-c-variation/05_regional_diversity.png)

*Figure 5. Mean position-specific non-S1133 frequency by region. The adjacent full-length AlphaFold trimer was rendered in PyMOL as a molecular surface with a mesh overlay. The structural DE-loop region, aa 250 to 273, was shown in dark blue within the globular head. Region boundaries followed the experimentally resolved 2JJL structure where available. Structural location did not establish receptor contact, antigenicity, or functional effect.*

### Cluster signatures and within-cluster polymorphism were not the same signal

A site that differed from S1133 in nearly every member of one cluster could be a stable lineage signature. It did not necessarily represent recent change or ongoing diversification. In contrast, several substantial residue states within the same cluster indicated within-cluster polymorphism, mixed sampled populations, or other evolutionary processes that required fuller phylogenetic analysis.

This distinction is essential when reading the animation. Red areas show frequent differences from S1133 within the active cluster. They do not show when those differences arose, whether they were selected, or whether they altered phenotype.

![Figure 6A. GC1 top and longitudinal views.](/images/molecular-biology/avian-reovirus-sigma-c-variation/06a_gc1_top_and_longitudinal.png)

![Figure 6B. GC2 top and longitudinal views.](/images/molecular-biology/avian-reovirus-sigma-c-variation/06b_gc2_top_and_longitudinal.png)

![Figure 6C. GC3 top and longitudinal views.](/images/molecular-biology/avian-reovirus-sigma-c-variation/06c_gc3_top_and_longitudinal.png)

![Figure 6D. GC4 top and longitudinal views.](/images/molecular-biology/avian-reovirus-sigma-c-variation/06d_gc4_top_and_longitudinal.png)

![Figure 6E. GC5 top and longitudinal views.](/images/molecular-biology/avian-reovirus-sigma-c-variation/06e_gc5_top_and_longitudinal.png)

![Figure 6F. GC6 top and longitudinal views.](/images/molecular-biology/avian-reovirus-sigma-c-variation/06f_gc6_top_and_longitudinal.png)

![Figure 6G. GC7 top and longitudinal views.](/images/molecular-biology/avian-reovirus-sigma-c-variation/06g_gc7_top_and_longitudinal.png)

*Figure 6A to 6G. Top and longitudinal surface views of the full-length Sigma C trimer for GC1 through GC7. Each panel used the same reference-based color scale and reported the number of sequences assigned to that cluster. Gray indicated no observed difference from S1133; progressively darker color indicated a higher within-cluster difference frequency. The views showed spatial patterns, not mutation timing, selection, or functional effect.*

### A candidate antigenic peptide varied, but one residue remained highly conserved

Goldenberg and colleagues predicted several Sigma C antigenic peptides and tested the S1133 peptide `THIPSDL` by ELISA. Antiserum raised against S1133 whole virus recognized that peptide more strongly than antiserum against a divergent isolate.[3] The experiment supported serum recognition of a linear peptide. It did not establish a neutralizing epitope or receptor-contact site.

In the local alignment, `THIPSDL` mapped to S1133 residues 254 to 260 within the globular head. Its consensus sequence differed among the seven clusters. D259, however, remained nearly invariant. That conservation makes D259 a reasonable candidate for targeted study, but it cannot distinguish among receptor contact, structural constraint, maintenance of loop geometry, or shared ancestry.

![Figure 7. Cluster consensus across THIPSDL.](/images/molecular-biology/avian-reovirus-sigma-c-variation/06_candidate_peptide_consensus.png)

*Figure 7. Cluster consensus residues at positions 254 to 260. Pink cells differed from S1133, while the gold column marked D259. Consensus conservation did not establish receptor contact or neutralizing function.*

## Interpretation

The most important result was spatial rather than numerical: sequence differences were not confined to the receptor-binding head. They extended along the scaffold that holds and presents it.

This observation supports several testable ideas. Nonconservative stem substitutions could alter local electrostatics, interprotomer packing, or flexibility. Changes near the head could alter surface chemistry or antibody recognition. A stem change could also influence how a chemically unchanged head was oriented or exposed. None of these mechanisms was demonstrated by the sequence map.

The head should remain the first priority for attachment and neutralization work because its receptor-binding role has direct experimental support.[1] The stem should be studied in parallel because it forms the trimeric architecture that makes the head functional as a fibre.[2]

## Implications for surveillance and vaccine matching

Genetic-cluster labels provide useful shorthand, and published surveillance studies have used Sigma C phylogeny to organize field diversity.[4,5] A label alone, however, does not describe every exposed surface or every difference within a cluster.

For autogenous vaccine selection, the map supports a staged workflow:

1. Sequence contemporary field isolates across near-full-length Sigma C.
2. Place them within a transparent, reference-anchored framework.
3. Compare exposed head patterns and structurally notable stem differences among candidate isolates.
4. Select a small, nonredundant panel representing the diversity relevant to the production system.
5. Validate the selection with matched serum neutralization and, where appropriate, controlled challenge studies.

Sequence similarity can prioritize isolates. It cannot replace antigenic or protection data. Vaccine performance may also depend on other viral proteins, reassortment, formulation, antigen integrity, administration, host immunity, and field exposure.

## Limitations

- Public NCBI sequences were unevenly sampled across years, countries, hosts, laboratories, and outbreak settings.
- Exact-sequence deduplication reduced deposit-level redundancy but did not remove broader sampling bias.
- Reference-anchored p-distance assignment was an operational classification, not formal taxonomy.
- The BIONJ presentation tree had no branch-support analysis and should not be used for strong evolutionary claims.
- The full-length trimer used for mapping was an AlphaFold model. Experimental Sigma C structures covered residues 151 to 326 and 117 to 326, not the complete 326-residue fibre.[1,2]
- Difference frequency relative to S1133 did not measure mutation rate, selection, immune escape, pathogenicity, or vaccine failure.

## Experimental priorities

The sequence map provides a rational starting point for experiments that could establish function:

- Introduce selected head and stem substitutions into trimeric Sigma C constructs or matched viral backgrounds.
- Confirm expression, folding, and oligomerization before interpreting cell-binding results.
- Measure trimer stability by orthogonal biochemical methods, including size-exclusion, native-gel, thermal-shift, and protease-sensitivity assays.
- Test cell attachment and neutralization with matched mutants and contemporary field sera.
- Compare representative clusters in controlled challenge studies only after sequence and antigenic panels have narrowed the candidates.
- Rebuild the cluster analysis with a model-selected maximum-likelihood method, branch support, and a preregistered GC7 anchor set.

![Figure 8. Experimental priorities for the Sigma C stem hypothesis.](/images/molecular-biology/avian-reovirus-sigma-c-variation/08_experimental_priorities.png)

*Figure 8. The sequence and structure map identified an experimental sequence: verify properly folded trimeric material or matched viral backgrounds first, then test stability, attachment, neutralization, and protection.*

## Conclusion

The globular head remains central to Sigma C biology, but it is not the whole story. Across 904 reference-anchored near-full-length sequences, amino acid differences from S1133 extended through both the head and the stem. The broad NCBI comparison reinforced that result, with especially high diversity in the N-terminal and coiled-coil portions of the fibre.

The correct conclusion is not that stem substitutions cause vaccine escape or alter attachment. It is that the stem contains enough structured, recurrent diversity to deserve direct testing. Putting phylogeny beside structure made that experimental gap visible.

## Related resources

- [Avian reovirus reference](/knowledge/viruses/avian-viruses/avian-reovirus)
- [Interactive protein structure viewer](/tools/protein-structure-viewer)
- [Molecular Biology knowledge base](/knowledge/molecular-biology)
- [X-ray crystallography guide](/knowledge/molecular-biology/x-ray-crystallography)

## References

[1] Guardado-Calvo P, Fox GC, Hermo Parrado XL, et al. Structure of the carboxy-terminal receptor-binding domain of avian reovirus fibre Sigma C. *Journal of Molecular Biology*. 2005;354(1):137-149. [https://doi.org/10.1016/j.jmb.2005.09.034](https://doi.org/10.1016/j.jmb.2005.09.034)

[2] Guardado-Calvo P, Fox GC, Llamas-Saiz AL, van Raaij MJ. Crystallographic structure of the alpha-helical triple coiled-coil domain of avian reovirus S1133 fibre. *Journal of General Virology*. 2009;90(3):672-677. [https://doi.org/10.1099/vir.0.008276-0](https://doi.org/10.1099/vir.0.008276-0)

[3] Goldenberg D, Pasmanik-Chor M, Pirak M, et al. Genetic and antigenic characterization of Sigma C protein from avian reovirus. *Avian Pathology*. 2010;39(3):189-199. [https://doi.org/10.1080/03079457.2010.480969](https://doi.org/10.1080/03079457.2010.480969)

[4] Egaña-Labrin S, Hauck R, Figueroa A, et al. Genotypic characterization of emerging avian reovirus genetic variants in California. *Scientific Reports*. 2019;9:9351. [https://doi.org/10.1038/s41598-019-45494-4](https://doi.org/10.1038/s41598-019-45494-4)

[5] Bilal AR, Jude R, Crossley B, et al. Surveillance of avian reoviruses in a single broiler chicken company from the California Central Valley. *Avian Diseases*. 2025;69(1):58-64. Published online December 6, 2024. [https://doi.org/10.1637/aviandiseases-D-24-00045](https://doi.org/10.1637/aviandiseases-D-24-00045)