Can immunohistochemistry reliably localize SARS-CoV-2 spike protein in tissue biopsies and how are controls handled?
Executive summary
Immunohistochemistry (IHC) can localize SARS‑CoV‑2 spike protein in formalin‑fixed, paraffin‑embedded (FFPE) tissues when rigorously validated antibodies and strict controls are used, but its reliability collapses in specimens with low viral burden and without orthogonal confirmation; several groups report both sensitive, specific protocols and widespread false‑positive/nonspecific staining with untested reagents [1] [2] [3]. Standardization—antibody validation on transfected cell pellets, dilution/titration windows, serial sections compared with RNA in situ hybridization (ISH) or PCR, and inclusion of infected and uninfected controls—is essential to interpret spike staining correctly [4] [5] [6].
1. IHC works — but only when antibodies are vetted and protocols optimized
Multiple teams identified commercially available antibodies that, under optimized conditions, produce clear, reproducible IHC detection of spike (and nucleocapsid) in FFPE lung and placental tissues from COVID‑19 patients, demonstrating sensitive and specific chromogenic labelling in high‑viral‑load samples [1] [7] [8]. Experimental work using cell pellets expressing defined viral proteins or pseudotyped virus has been used to confirm antibody specificity and staining patterns, showing that some monoclonals detect SARS‑CoV‑2 spike specifically while not labelling MERS or untransfected controls [4] [2].
2. The cautionary counterpoint: nonspecific staining and low‑load samples
A multicenter assessment warned that in tissues with low viral RNA loads IHC is not reliable to determine organ tropism or specific infected cell types because rare immunosignals may be misread and nonspecific background interpreted as true positives; anti‑spike antibodies in particular may give discordant results if used without strict validation [3]. Case series also report negative spike IHC despite detectable subgenomic RNA or PCR in tissues like retina/optic nerve, underscoring that negative IHC does not exclude RNA presence and that IHC may fail when antigen abundance is low or degraded [9] [10].
3. How controls are—and should be—handled in robust studies
Best practices documented across studies include multiple, complementary controls: positive controls (known COVID‑19 tissues, infected cell pellets or transfected HEK293 pellets), negative controls (pre‑pandemic tissues, untransfected cell pellets), antibody dilution series to find the signal‑to‑background window, and concordant testing on serial sections by RNA ISH or RT‑qPCR; blinded reads and concurrent testing of multiple viral proteins improve confidence [2] [5] [6] [7]. Standardization protocols explicitly recommend sequentially labeled serial sections, testing multiple antibodies and viral targets, and documenting no background at optimized dilutions (for example, spike S2 showing no background at very high dilutions in one protocol) [5].
4. Interpreting discordance: spike protein without RNA and vice versa
Some dermatopathology and autopsy reports describe IHC spike positivity in tissues where RNA is undetectable, raising plausible biological explanations (circulating or cleaved spike protein deposited in endothelia) but also flagging methodological pitfalls if orthogonal RNA testing isn’t performed; conversely, detection of viral RNA without spike IHC has been observed, suggesting limits of antigen preservation, assay sensitivity, or timing in infection [11] [9] [10]. These discordances mean IHC alone should not be the sole arbiter of tissue infection; rather it must be interpreted in the context of RNA assays, clinical timing, and assay validation [5] [3].
5. Bottom line and competing agendas in the literature
The literature converges on a conditional verdict: IHC can reliably localize spike protein when using validated antibodies, appropriate positive/negative and procedural controls, and orthogonal confirmation, but many early and some later studies used unvalidated reagents or insufficient controls and produced misleading claims about multi‑organ tropism [1] [2] [3] [7]. Authors promoting single‑assay proofs of widespread tissue infection may be biased by limited assays or urgent publication pressures, while standardization-focused groups explicitly call for multicenter protocols and cross‑validation to avoid overinterpretation [5] [3].