The Missing Receptor: What Negative Proteomics and Annotation Evidence Says About Bovine GPRC6A

The story of bovine GPRC6A is not a simple courtroom drama where one Western blot wins and one database entry loses. It is stranger and more interesting. On one side, several bovine studies report GPRC6A protein detection, siRNA knockdown behavior, receptor localization, and nutrient-linked signaling in mammary epithelial cells. On the other side, large proteomic atlases, broad transcriptomic searches, and major annotation databases do not consistently see a clean, canonical bovine GPRC6A protein.

This blog post is about the second side: the negative evidence.

Not negative in the sense of “GPRC6A cannot exist in cow.” That would be too strong. Rather, negative in the more careful sense:

If bovine GPRC6A were a broadly expressed, abundant, canonical full-length protein, some large-scale resources should probably have detected it more clearly. Many did not.

That silence matters.

1. VPBrowse: the strongest broad bovine proteomics non-detection

The most important negative source is:

Paramasivan S, Ashick M, Dudley KJ, Satake N, Mills PC, Sadowski P, Nagaraj SH. 2024. “VPBrowse: Genome-based representation of MS/MS spectra to quantify 10,000 bovine proteins.” PROTEOMICS 24(14):e2300431. DOI: 10.1002/pmic.202300431.

VPBrowse is not a small experiment. It is a broad bovine spectral-library resource. The authors describe it as an online platform for genome-based representation of the Bos taurus proteome, designed for searching, visualizing, and building quantitative mass-spectrometry assays. The publication details show that the article was published in PROTEOMICS volume 24, issue 14, article e2300431.

The scale is the reason VPBrowse matters. The paper reports a library of 10,500 proteins, with 36,064 peptides mapped to 10,255 proteins, supporting quantification of about 27% of the Bos taurus UniProtKB protein space. It also says that 79% of proteins had at least two proteotypic peptides, and Figure 1 summarizes the draft cattle proteome, tissue/body-fluid distribution, chromosomal coverage, and peptide-per-protein depth.

So this is exactly the kind of resource where one might expect a detectable bovine GPRC6A peptide to appear, especially if the canonical protein were broadly detectable across tissues or fluids.

The negative claim comes from the 2026 reanalysis:

Soman AS, Vijay N. 2026. “Re-evaluating GPRC6A Gene Loss: Implications for Milk Fat Synthesis and Functional Genomics in Dairy Cattle.” Cureus Journal of Agriculture and Food Science 2:es44497-026-00164-y. DOI: 10.7759/s44497-026-00164-y.

That reanalysis states that searches of the Veterinary Proteome Browser did not identify peptides corresponding to GPRC6A. It also frames the whole GPRC6A problem as a discrepancy between experimental functional evidence and sequence-based gene-loss predictions.

This is the strongest negative proteomics evidence currently in the stack.

But it still needs careful interpretation. VPBrowse non-detection does not prove that GPRC6A protein is absent in every cow tissue or condition. GPRC6A is a GPCR and a membrane protein. Membrane proteins are often under-represented in proteomics because their hydrophobicity, low abundance, poor solubility, and digestion/extraction behavior make them hard to identify at scale.

So the fair conclusion is:

VPBrowse strongly argues that bovine GPRC6A is not broadly or easily detected in large-scale bovine proteomics. It does not rule out low-abundance, tissue-specific, condition-specific, truncated, or isoform-specific protein expression.

2. Bovine PeptideAtlas: mammary-relevant but probably negative

The second important source is:

Bislev SL, Deutsch EW, Sun Z, Farrah T, Aebersold R, Moritz RL, Bendixen E, Codrea MC. 2012. “A Bovine PeptideAtlas of milk and mammary gland proteomes.” PROTEOMICS 12(18):2895-2899. DOI: 10.1002/pmic.201200057.

This source is especially relevant because it focuses on milk and mammary-gland biology. The Bovine PeptideAtlas was built as a resource for selected-reaction-monitoring assay design for milk production and mammary gland health. It includes 107 samples from six tissues, with 1,921 canonical proteins at 1.2% FDR and 8,559 distinct peptides at 0.29% FDR.

The tissue coverage is also highly relevant. The atlas includes mammary epithelial cells, colostrum, milk, udder tissue, hoof, and a mitochondrial fraction. Mammary epithelial cells, colostrum, milk, and udder tissue are precisely the places where one might hope to see protein support for a receptor implicated in mammary biology.

The important caveat: I did not find a direct indexed GPRC6A/E1BPQ3/GPC6A hit in the publicly surfaced PeptideAtlas article text or search snippets. That makes it a likely negative, not a definitive negative.

Why weaker than VPBrowse? Because this atlas is older and shallower. It covers about 9% of predicted bovine proteins, according to the article, and was built with older genome/protein annotations.

Its Figure 1 is not a GPRC6A figure. It illustrates the PeptideAtlas protein-view concept, showing how users can inspect protein sequence coverage, observed peptides, predicted observable peptides, and supporting spectra.

The fair conclusion is:

The Bovine PeptideAtlas had mammary-relevant samples and peptide-level search infrastructure, but no public indexed GPRC6A hit surfaced in this search. This is a meaningful but moderate negative, not a final absence claim.

3. PXD031744: bovine milk top-down dataset, no public GPRC6A signal found

A third useful source is the ProteomeXchange dataset:

ProteomeXchange / MassIVE dataset PXD031744. “Top-down proteome of de-fatted Bos taurus milk from an animal with clinical mastitis.” Announced 2022. Repository: MassIVE. Primary submitter: David L. Tabb.

The dataset collected quarter-milk samples from a cow with clinical mastitis on days 13 and 16 postpartum, comparing a clinically abnormal quarter with a normal quarter from the same animal. It used LC-MS on Orbitrap instruments and is explicitly labeled as Bos taurus milk.

This dataset is relevant because it samples milk, a mammary output, under inflammatory/stress conditions that could plausibly alter low-abundance proteins. It is not a mammary epithelial membrane-enriched dataset, so the expectation is weaker than for a targeted membrane proteome.

The associated top-down proteoform benchmarking paper is:

Kou Q, et al. 2023. “Comparing Top-Down Proteoform Identification: Deconvolution, PrSM Overlap, and PTM Detection.” Journal of Proteome Research 22. DOI: 10.1021/acs.jproteome.2c00673.

That paper used bovine milk data from PXD031744 while benchmarking top-down proteoform identification and PTM detection. Public search results did not surface a GPRC6A/E1BPQ3/GPC6A_BOVIN hit for PXD031744.

This is a weaker negative because milk is dominated by abundant secreted proteins, and top-down milk proteomics is not optimized for low-abundance GPCRs. Still, the absence is useful as a boundary marker.

The fair conclusion:

PXD031744 is a weak-to-moderate negative. It had some chance to detect unusual bovine milk proteins, but failure to find GPRC6A there is not surprising and cannot outweigh targeted BMEC protein evidence.

4. Broad RNA-seq reanalysis: no clean full-exon expression pattern

The negative picture is not only proteomic. It also appears in transcriptomic reanalysis.

The 2026 Soman and Vijay reanalysis reports that all six bovine GPRC6A exons were queried across a MetaGraph-indexed Bos taurus RNA-seq database containing many public SRA accessions. According to the article, no individual RNA-seq sample showed expression across all GPRC6A exons, and the highest-abundance sample showed reads only in exons 1, 5, and 6, not exons 2, 3, and 4.

This does not directly answer the protein question. RNA absence is not protein absence, and fragmented transcript evidence could reflect annotation problems, tissue specificity, low expression, or partial transcripts.

But it does add pressure against a simple story of abundant full-length canonical GPRC6A expression.

The fair conclusion:

Broad RNA-seq evidence, as summarized by the reanalysis, supports skepticism about a clean full-length canonical bovine GPRC6A transcript, while leaving open shorter, tissue-specific, or condition-specific transcripts.

5. Gupta et al. 2026: comparative genomics argues for Bovidae-wide gene erosion

The most important genomic skepticism source is:

Gupta S, Patil AB, Soman AS, Vijay N. 2026. “Master of none: GPRC6A gene loss is more widespread than previously known.” Genetica 154(1):5. DOI: 10.1007/s10709-026-00258-7.

This paper is not a proteomics paper. It is a comparative genomics study. Still, it matters because it asks whether the bovine GPRC6A locus is intact in the first place.

The PubMed abstract says the authors used a synteny-informed comparative genomic approach and concluded that GPRC6A loss is more widespread than previously reported, including the entire Bovidae group within Artiodactyla.

This creates the central tension.

If a comparative-genomics paper says the locus is disrupted across Bovidae, but BMEC studies report GPRC6A protein and function, then one of several things must be true:

The proteomics and Western blot signals may reflect a truncated or alternative product.
The annotation may miss an unresolved genomic copy.
The functional studies may detect a related protein or antibody cross-reactivity.
The genome-loss inference may be too binary for this locus.
Some mixture of these may be happening.

The fair conclusion:

Gupta et al. is not a negative proteomics source, but it is strong genomic evidence that bovine GPRC6A should be treated as an unresolved locus rather than a routine intact protein-coding gene.

6. NCBI Gene: curatorial warning from the reference annotation

NCBI Gene currently lists:

GPRC6A, G protein-coupled receptor class C group 6 member A, Bos taurus, Gene ID 783669.

The important details are that NCBI lists the bovine gene type as pseudo and the RefSeq status as INFERRED. It still links model transcript/protein accessions and UniProtKB/Swiss-Prot E1BPQ3, but the pseudo label is a major warning sign.

This is not a failed proteomics experiment. It is not a direct absence claim. But curatorial databases integrate genome structure, annotation evidence, transcript/protein models, and comparative data. If NCBI marks the gene as pseudo, that means the canonical gene model is not being treated as a straightforward experimentally supported protein-coding gene.

The fair conclusion:

NCBI Gene supports annotation-level skepticism: bovine GPRC6A is not currently a clean, settled protein-coding annotation in RefSeq.

7. UniProt E1BPQ3: the protein exists as an entry, but not with accepted direct protein-level evidence

UniProt has the bovine entry:

E1BPQ3 / GPC6A_BOVIN, “G protein-coupled receptor family C group 6 member A,” Bos taurus.

The entry gives a complete 888-aa sequence and includes the motif NDVFIVTNQETK, which is important because it has been reported as a candidate bovine GPRC6A peptide in later reanalysis.

But UniProt’s evidence level remains “Inferred from homology”, not evidence at protein level. UniProt’s help explains that “inferred from homology” means the existence of a protein is considered probable because clear orthologs exist in related species.

That is a subtle but important negative.

UniProt is not saying “no protein exists.” It is saying that, as a curated protein entry, bovine GPRC6A has not been promoted to direct protein-level evidence.

The fair conclusion:

UniProt preserves a bovine GPRC6A protein model, but it does not yet treat the protein as experimentally confirmed at the highest evidence tier.

What the negative evidence means when placed beside the positive evidence

This is where the story becomes interesting.

The negative evidence says:

Bovine GPRC6A is not broadly visible as an ordinary, abundant, canonical protein across general bovine proteomics and annotation resources.

But positive evidence does exist.

Singh et al. 2018 list E1BPQ3, annotated as “G protein-coupled receptor family C group 6 member A,” Bos taurus, GN = GPRC6A, in crossbred bull spermatozoa Table 3, with PLG score 52.72 and a good-vs-poor ratio of 0.794534.

Li et al. 2019 report that GPRC6A knockdown “totally abolished” lysine-stimulated PI3K phosphorylation and downstream FABP5/SREBP-1c signaling in bovine mammary epithelial cells.

Jin et al. 2022 report that GPRC6A knockdown “almost totally blocked” palmitic-acid stimulation of PI3K and PKCα activation, and reduced downstream SREBP-1c and triglyceride outputs.

Yu et al. 2019 are especially useful as an honesty control: GPRC6A knockdown did not suppress taurine signaling, but the paper still used GPRC6A antibody and GPRC6A siRNA in a receptor-exclusion experiment.

Zhang et al. 2025 add an in vivo cow layer, reporting that mammary GPRC6A, p-Akt/Akt, and p-mTOR/mTOR increased with dietary rumen-protected arginine, and concluding that RPArg activates the GPRC6A-Akt/mTOR pathway.

So the positive and negative evidence do not cancel each other neatly. They create a sharper hypothesis.

The careful synthesis

A weak synthesis would be:

“Some papers say GPRC6A exists, others say it does not.”

That is too crude.

A better synthesis is:

Bovine GPRC6A is not well supported as a broadly expressed, canonical full-length protein across large-scale public proteome and transcriptome resources. However, several targeted bovine studies report protein detection, siRNA-sensitive bands, receptor localization, and pathway effects in specific cell types and conditions. Therefore, the strongest current model is not simple presence or absence, but conditional, low-abundance, tissue-specific, or noncanonical GPRC6A protein biology.

In other words, the negative evidence changes the claim from:

“Cow GPRC6A is obviously intact and functional everywhere.”

to:

“Cow GPRC6A has targeted experimental support, but broad proteomics, transcriptomics, and annotation resources suggest the locus/product is unusual and needs careful validation.”

That is the honest landing place.

Ranking the negative evidence

Here is the practical ranking:

Source	Negative signal	Strength
VPBrowse / PXD044768	Deep bovine spectral library, reported no GPRC6A peptides	Strong
Gupta et al. 2026	Comparative genomics argues Bovidae-wide GPRC6A loss/erosion	Strong genomic skepticism
NCBI Gene	Gene type listed as pseudo, RefSeq inferred	Strong annotation skepticism
UniProt E1BPQ3	Protein entry remains inferred from homology	Moderate curatorial skepticism
Bovine PeptideAtlas	Mammary/milk atlas, no indexed GPRC6A hit found	Moderate but not definitive
PXD031744	Bovine milk top-down dataset, no public GPRC6A hit found	Weak to moderate
MetaGraph/SRA reanalysis	No clean all-exon transcript sample reported	Useful transcriptomic skepticism

Final conclusion: a quiet absence, not a guillotine

The expanded negative evidence is real. VPBrowse in particular is a serious missing signal. Bovine PeptideAtlas and PXD031744 add smaller missing signals from mammary/milk proteomics. NCBI and UniProt show that curators remain cautious. Gupta et al. adds the genomic thundercloud: the locus may be disrupted across Bovidae.

But the negative evidence is not enough to erase targeted experimental studies. For a GPCR, especially a low-abundance membrane receptor, broad discovery proteomics can miss true proteins. The right conclusion is more interesting than “present” or “absent.”

Bovine GPRC6A should be treated as a disputed, noncanonical, possibly condition-specific protein system.

The next decisive experiments would be:

Targeted PRM/MRM mass spectrometry for unique bovine GPRC6A peptides.
Immunoprecipitation followed by MS to verify the Western blot band.
Long-read RNA-seq from BMECs, testis, sperm, mammary gland, and buffalo/cow Leydig cells.
CRISPR knockout or precise exon-targeted perturbation in BMECs.
Rescue experiments with the bovine predicted ORF or candidate truncated ORFs.

Until then, the honest blog headline is this:

Large-scale resources often fail to see bovine GPRC6A, but targeted biology keeps leaving footprints. The cow receptor is not a clean textbook gene. It is a foggy little signal that needs proteogenomic daylight.